Mod 5-Fan Measurement and Testing
Mod 5-Fan Measurement and Testing
Training Program
MODULE 5
Fan Measurement and Testing
Section Content
4 Fans, FL Smidth
6 Fan Applications
PRESENTATIONS
Bowmanville Calciner ID Fan Problem
Fan Technology
Centrifugal Fans and Systems
Blue Circle Cement
Module 5
Section 1
Page No.
Fan total pressure PT, is the difference between the total pressures at the fan inlet and outlet.
Fan static Pressure, PS,is the fan total Pressure minus the fan velocity pressure.
u2
Fan, velocity pressure, PV, is the velocity pressure corresponding to the average velocity at the fan outlet.
2g
i.e. the fan static -pressure is NOT equal to the difference between the static pressures at the inlet Ar!-.1 outlet.
It should be measured as the difference between the reading of a facing tube at the inlet and a side tube at the fan
outlet (see Figure 1).
These show the effect on fan pressure, Power and efficiency of a change in fan throughput (Fig.2). The
horse-power on a fan curve is that which must be delivered to the shaft and does not take into account losses in
couplings and motors.
A calculation of the work actually done on the air will therefore give the efficiency of the fan itself,
5.2
Work done on air = pressure x flowrate x HP
33000
PV
= HP (inches wg and cfm)
6360
P may be either the fan static or total pressure and the efficiencies thus calculated are the static and total
efficiencies respectively.
Because work is being done on the air, its internal energy will rise, and this, together with losses due to the
inefficiency of the fan will cause the air temperature to rise.
fan Power
Temperature rise = (V )(specific heat )
density
PV × 5.2
779 × η
For air at ambient conditions = η = total fan efficiency
0.075 × V × 0.24
P o
= 0.37 F
η
o
or about ½ F per inch wg.
The resistance of the system served by a fan depends upon the hardware in it and for any system there will be a
relationship between flowrate and pressure drop. If the system contains only dampers, ducting, expansions etc.,
of fixed resistance, then the pressure drop may be taken as being proportional to V2 (approximately). This would
not be true in the case of a resistance (as opposed to pressure drop), that varied with flowrate, such as a
Lepol grate. The intersection of the system resistance line with the fan pressure/flowrate curve gives the
point of operation of the fan. The effect of, for instance, closing a damper is to steepen the system resistance
line, giving a new operating point (see Fig.3), System resistance lines may be calculated from fluid flow theory
or one point may be measured end the remainder drawn in assuming, for instance, a squared relationship.
2.1 The Shape of Theoretical and Practical Fan Curves Fan curves may be calculated theoretically and the
shape of these curves are shown in Fig. 4. However, in practice losses modify the curves to give what we
normally recognize as a typical fan curve.
Friction at the fan inlet, casing, and blade passagescauses a loss which is proportional to the square
of the velocity (or volume) and is zero at zero flow.
FIG. 4 FAN CURVES CALCULATED FROM THEORY
These occur at the blade leading edge and in the fan casing. Depending upon the relative velocity
between the air and the impeller leading edge, the angle of attack will change. The smoothest flow
will occur when the above relative velocity is in the same direction as the blades themselves. Here
eddy loss is zero and increases as the angle of attack moves away from zero.
A similar situation occurs in the fan casing, there being only one flow where eddies will not occur. It is usual to
design casings such that the flowrate corresponding to eddyless flow is the same for both impeller and casing.
(c) Leakage:
Backflow between casing inlet and impeller inlet causes a small loss.
(e) The theoretical no-lose characteristics shown above assume that the air follows the blade passages
absolutely. This only be true for an infinite number of blades.
The effect of the above on fan characteristics is shown in Figures 5A and. 5B for a backward curved impeller.
The performance of a fan may be changed by several methods, the main ones being a, change in
a) fluid density, p
c) fan size * Most text books only quote the fan laws for "fans of geometric similarity", so that
diameter, width, outlet duct size, etc., are all increased at the same rate. Any convenient dimension
may then be taken as being representative of the fan, e.g. Impeller Diameter, Dim,
ALTERATION IN
CHANGE IN FAN VOLUME POWER
PRESSURE
Density, p αp No change αp
Speed, n αn 2 αn α n3
Fan Size*, Dim α Dim2 α Dim3 α Dim5
Fan pressure may be either total, static or velocity. In order to help remember the above table, it is worth noting
that fan pressure can be shown to be theoretically proportional to the square of the tip speed whereas volume is
proportional to the swept volume of the impeller (fan size/volume is a cubic relationship since width increases
we well as diameter). Power is proportional to the product of volume and pressure.
A change in impeller diameter/width ratio will change the characteristic of the fan and direct scale-up is not
possible, although if a fan is running at somewhere near its peak efficiency (i.e. just to the right of the peak on
the pressure - Volume graph) then for a small increase in diameter, the volume will increase as the cube and the
pressure as the square of the tip speed, i.e. the fan laws in Table I would appear to apply. However, it must be
remembered that a new fan type has been produced by the change in diameter/width ratio and throttling of the
fan will not follow any easily calculable curve.
Presented with curves for a fan running at any set of conditions, it should be possible to predict the curves at any
other set of conditions.
Before setting out on the calculations, it is worth checking that the fan is actually operating close to the
manufacturer's fan curve. Their fan tests are performed with straight inlet and outlet ducts and, for instance, a
bend close to the fan inlet on site could lead to adverse whirl and a reduction in the fan efficiency. The method
of calculation is best shown by example. Appendix I shows the effect of a change in density and speed on a fan,
these being the variables normally encountered in practice.
Care must be taken in applying the fan laws, they are only applicable over a limited range because they are only
an approximation. A better estimate is given by
The indicies x and z are so small, however, that the effect of the Reynolds Number is usually negligible in
practical cases.
Even British Standard 848 states that Reynolds Numbers may be ignored in scale up provided they do not vary
by more than 40%.
1. Damper Control
2. Variable Radial Inlet Vanes - these give a beneficial .swirl to the air at the fan inlet and therefore reduce the
work the fan has to do.
3. Speed Control.
Initially, dampers are the cheapest to install and speed control the most expensive, but there is a reversal in a
running cost comparison. - A cost comparison must be made in each individual case based on the capital cost,
running cost, and the amount of turndown that will be needed. However, speed control usually only becomes
economic on the larger machines.
Whereas variation in a damper setting will change the system resistance and not the fan curve, the other two
methods produce new fan curves, that for speed control retaining the same fan efficiency. Since the use of speed
control introduces the inefficiency of the device itself, it is worth noting that for small changes from the
maximum throughput, speed control has, in fact, the highest running cost.
Referring to Figure 6, the effect of turndown to 3 V/4 and V/2 by the three methods is shown.
Graph A - Damper Control - Fan curve retained but system resistance line altered.
Graph B - Variable Inlet Vanes - New fan curves produced for each vane setting.
Graph C - Speed Control - New fan curves produced for each speed but efficiency retained.
This graph assumes no pressure drop across a fully open damper. This may be a reasonable assumption for some
dampers operating in clean air, but would certainly not be true for some I.D. fan dampers where build-up will
occur. The expected pressure drop across an open damper, based on experienced, should be taken into account
when deciding on the method of control.
Radial inlet vanes are not cheap to install and can increase the cost of a fan by 20%. However, they are an
efficient means of control over a limited range and should be considered as an alternative to speed control. For
larger variations in volume a two or three speed motor with final adjustment by inlet vanes would be cheaper
than infinitely variable speed control.
If considerable operation at less than maximum capacity is expected, speed control had the added advantages of
an increase in life and a reduction in noise level.
Fan performance may be affected by worn impellers, inefficient drives, etc., but the most common
causes of fan performance being substandard are bends close to the inlet or outlet of the fan.
If the bend is at the inlet, the air is pushed over to one side of the eye, whereas a manufacturer's fan curve is for
uniform inlet conditions. Uniformity of flow at inlet is easily checked by a pitot traverse.
The best solution to this is an inlet box which can be sized so that the air enters the impeller uniformly, failing
that a square bend with cascade (Fig.7) will give good performance. Retention of the original bend and fitting of
long turning vanes will improve matters greatly (Fig.8).
Useful pressure in a fan originally comes from the acceleration of the air, - As the air is again slowed down by
expansion, velocity pressure is converted into static pressure. For this to be done efficiently the expansion
should be slow and gentle and clearly a bend immediately on the fan outlet will not help matters. This problem
is not so easy to remedy but again a cascade will help.
6.0 SERIES FANS
For fans in series, the second fan will handle -the same mass of gas as the first fan. However, due to
compression effects and temperature effects (Equation 3). The second fan will not handle the same volume. To
be absolutely accurate, the fan curve should therefore be modified to correct for the density difference. The two
effects, however, tend to cancel each other out, and in practice it is usually acceptable to use the original curve
for the second fan. The combined pressure at any volume is found by adding the total pressures of the individual
fans and subtracting the losses for the ducting between the fans (usually negligible) - see Fig. 9. Here to obtain
the complete fan curve it has been necessary to know the characteristic of one fan at volumes greater than it
would handle with free inlet and discharge (i.e. being "force-fed"). This is not normally known, and in any case
there would be little point in operating under these conditions since the fan would be hindering the flow of air
rather than helping it.
Unless the fans are identical it is likely that one of them will have to operate at a point that does not give its
maximum efficiency. For this reason, if a series set up is necessary, it is customary to install identical units. This
may be necessary if the space available renders one large fan impractical or if it is necessary to keep the air
flowing while one of the fans is being maintained (each fan would then have a by-pass).
A similar exercise is then performed on Fan 2 and its associated system. The two curves (P-F)1 and (P-F)2 are
then added together to give the characteristic of the parallel unit. Any system resistance line now drawn in a
refers only to the resistance of ducting, etc., that is common to both fans. It should be noted that alteration of any
damper in the non-common ducting will change the parallel unit characteristic rather than the external system
resistance line.
Not all fans have such simple characteristics as those in Fig. 10. For instance, parallel operation of two identical
forward curved blade fans is shown in Fig.12.
At the pressure Pv, there are three possible operating volumes for the single fan, namely Vx, Vy, and Vz. For the
single fan the system resistance line would determine the point of operation. However, these three points rive
rise to six possible points of operation for the parallel system. Three of them are simply 2Vx, 2Vy and 2Vz and
they form the main part of the characteristic for the parallel arrangement However, the three other possible
additions (Vx + Vy), (Vx + Vz ) and (Vy + Vz ) from off-shoots to the main curve.
It can be seen that, depending on the position of the system resistance line there will be one, two or three
possible operating points. For the system resistance line drawn in Fig.12 there will in fact be three possible
points of operation. The fans may settle down at any of these positions or oscillate between all three. C would be
the desired operating point, when both motors will develop the same load, whereas there will be progressively
more imbalance as the system moves to-point B and then point A. In practice both motors may cut out
immediately on start-up, the second fan because it is overloaded and the second fan because it then tries to run
at the point D.
Hagen's method is a rather neat way of predicting the above unsatisfactory performance of identical fans. The
system resistance and total pressure single fan curve are drawn and, for any total pressure, the magnitude of the
term [(system volume flow) - (maximum possible fan volume flow at that pressure)] is superimposed. If the line
(Eagen's line) thus formed crosses the fan curve at more than one point then operation will be unsatisfactory
(Fig. 13).
This can be effected by fitting balancing dampers to each fan as shown in Fig.14. The system resistance line, D,
crosses the original parallel fan curve in three places leading to unsatisfactory operation. The balancing damper
and original fan curves are combined to give the curve (P-R) for each fan (the balancing damper bring
considered part of the fan) and when the curve is drawn for the two fans (P-R) in parallel it can be seen that the
system resistance line cuts it at only one point. Hence with careful adjustment of the balancing dampers it
should be possible to run the two motors with balanced loads, with only a small loss in volume flow.
8.0 BLADE TYPES
Fig. 15 shows typical fan curves for each type. In general for a given impeller size, the pressure developed
increases with blade angle and higher volume flows can be obtained by use of a double inlet.
These fans give the highest pressure for a given size and speed. They are also used for high volume applications
such as ventilation, where their compact size and low speed (hence low noise) is an advantage. The low speed
also allows a lighter, less expensive construction to be used.
The blades are of short radial depth to reduce the inlet throttle and consequently more of them (30-60 compared
with 6-16 for other fans) are required to have an effect on the air. It can be seen (Fig.15) that a decrease in
restriction and hence an increase in airflow will cause an increase in power. The fan is said to be of the
overloading type and this effect may cause the motor to trip out.
These fans are unsuitable for dirty air, the narrowly spaced tips tending to bridge. Efficiencies are not greater
than 75%
These have their main application in handling high dust burdens. They are of simple construction and
therefore the wearing parts can be of substantial material or even replaceable. The impeller is
unshrouded but may have a back plate in the case of fibrous dust to prevent build-up on the shaft.
For corrosive applications, it is a simple matter to rubber-cover the impeller. Disadvantages of the
paddle fan are its low efficiency (about 605), high speed of rotation (noisy operation) and
overloading power characteristics.
These fans combine some of the ruggedness of the paddle fan with the higher efficiency and higher
pressure development of the forward curve blade. The impeller is self-cleaning to a large extent and
is used extensively for dusty flue gases. Blades themselves are not easily made removeable but
replaceable wear plates and nose pieces can be fitted if dust burdens are high. Efficiencies higher than
75% are obtainable with this fan and it is therefore commonly used where high dust burdens prevent
the use of the more efficient aerofoil or backward inclined bladed fans.
8.3 Backward Bladed Fans
In the main, these fans are used for clean air applications, although the use of a reinforced nose, or
leading edge, will allow low concentration of dust to be handled. Efficiencies up to 90% are
obtainable with fans having aerofoil section blades (Fig.16) and therefore the fan is preferred for
the larger machines.
For a given duty the fan must run at a higher tip speed and the higher efficiency may be partially offset by the
capital cost involved in producing accurate aerofoil blading of sufficient robustness to withstand the increased
rotational stress.
The high efficiencies are a result of laminar flow over the blade profile and the reduction of eddies leads to
quieter operation. 'Backward bladed fans are nonoverloading (see power curve on Fig.15).
Forward Curved
Backward Aerofoil
Backward. Inclined
Forward Curved Radial Tip
Paddle
The main source of fan noise is caused by the formation of shock eddies as the air passes through the blades of
the impeller and therefore quietness depends upon a smooth passage of the air. Fans generally operate most quietly
at their highest efficiency and dampering back may cause an increase in sound level. The outlet throat velocity does
not in itself control the noise level - the clearance between the impeller and the tongue being more important. The
rate of expansion of the air caused by the casing shape is also of importance. An expansion that is too rapid will
give a low outlet velocity but a higher noise level. The noise generated by the fan is transmitted to the casing both
upstream and downstream causing resonance in ducting. Fan noise can be reduced by speed reduction (if the fan is
already heavily dampered), duct and casing stiffening, or the fitting of inlet sound attenuators of the "Quietflow"
type. Appendix II gives a method of predicting octave band sound pressure levels from which a dB(A) rating can
be calculated.
10.0 OTHER PUMPING EQUIPMENT FOR GASES
Table 2 shows the applications of the various types of pumps available, though the figures should be taken as a
general guide only :
Apart from the centrifugal fan, which is by far the most common in our industry, use is also made of Crossflow
and Axial flow fans.
Here the air enters and leaves radially (see Fig.17). This fan is used when a very low fan pressure is required
Such as in the domestic fan heater and some hair driers.
FIGURE 17 - CROSSFLOW FAN
Here, as the name implies the air enters and leaves the fan axially. They are a development of the propeller fan
(e.g. Xpelair) which is only suitable for low pressures (i.e. free inlet and outlet). The modern axial fan, however,
is much more versatile and has the advantage that the motor may also be mounted on the axis of the duct giving
a very compact unit. The motor may be sited outside the gas stream by use of a bifurcated duct (Fig.18). The
axial flow fan is generally not used in dusty conditions since the blade tip clearance tends to bridge
These are the only machines capable of very high delivery pressures (100 pai). Their high efficiency must be
balanced against the lower capital cost of a rotary compressor.
High compression types include the sliding vane blower (Fig.19) and the Nash Hytor (Fig.20). The Hytor has an
elliptical casing, partially filled with fluid which is thrown to the outside of the casing by the impeller. The
spaces in vanes act as liquid pistons so that the air is sucked in, compressed and then discharged by the
advancing interface. The liquid may be chosen to be inert to the gas being pumped.
Neither of the above two machines are affected by tip wear. The sliding vane compressor has loose blades which
simply move radially outwards as they wear whilst it can be seen that the Hytor will function provided the blade
tips are always covered by liquid. For this reason these machines will maintain their efficiency over long
periods. Typical of the lower compression ratio type is the Rootes Cycloidal Blower (Fig.21 ), where
Increased wear and therefore back-leakage will reduce efficiency.
These machines are used at anything from ½ psi to 100 psi and operate at high speeds (3500 to 30,000 rpm).
Water cooling may be employed in-between the stages and in-between casings which normally contain 6 or 7
stages.
The axial flow compressor is used for special applications (e.g. jet engines) where its lightness is an advantage.
It is a high efficiency machine but has a limited operating range, greater vulnerability to erosion and 00=03ion
and susceptibility to deposits.
Fig. 23 shows an axial flow compressor. The rotating element c6nsists of a single drum to which are attached
several rows off aerofoil blades, of decreasing height. Between each row is a stationary row which straightens
the airflow before it meets the next moving set of blades.
10.6 Vacuum Pumps
Here, a high compression ratio is necessary and therefore the machines used include the reciprocating piston,
rotary vane (Fig.24) and Nash Hytor, Piston and rotary pumps will give pressures down to 10-2 mm. Hg. but the
Hytor is limited by the vapor pressure of the sealing liquid (usually water).
11.0 POWER REQUIRED FOR COMPRESSION
This depends on the method of compression, which can be isothermal, adiabatic (no heat removed from
system)or something in-between the two. Isothermal compression requires the least work and is therefore the
most efficient so that the isothermal line on a P-V diagram should be approached as closely as possible, which
means that the maximum cooling between stages should be used.
P
Wi = P1 V1 log e 2
P1
For adiabatic compression n = γ = ratio of specific heats. Compressor efficiencies (based on the
isothermal power required vary between 45 (low compression rotary compressors)
and 60% reciprocating piston compressor)).
Most fan manufacturers have published their own technical books, the most comprehensive being
Given the fan curves shown at 50oC and 200 rpm what are the fan curves for running conditions of 0 oC and
220 rpm.
Step 1
220
Speed change factor = = 1.10
200
323
Density change factor = = 1.18
273
2
n p2
Pressure factor = 2 = (1.10)2 × 1.18
n1 p1
= 1.43
n
Volume factor = 2 = 1.10
n1
3
n p2
HP factor = 2 = 1.57
n1 p1
Step 2
Pick out reasonably spaced points on the original fan curves as follows:-
Volume
(cfm x 1000) 0 10 20 30 40 50 60 70 80 90 100
Fan Total
Pressure
(ins Wg) 9.7 9.8 9.9 10.0 9.7 9.2 8.2 6.8 5.1 3.4 1.5
HP 45 50 62 76 87 96 99 99 97 93 80
Step 3
Multiply by the appropriate factors to obtain the new fan curves:-
Volume 0 11 22 33 44 55 66 77 88 99 110
PT 13.9 14.0 14.2 14.3 13.9 13.1 11.7 9.7 7.4 5.0 2.1
HP 71 78 97 119 137 149 155 155 152 146 126
APPENDIX II
PWLs = Specific power level in each octave band (re 10-12 watt)
5 dB should be added to the octave band containing the blade passage frequency to account for rotational noise.
Water injection into an air stream before a fan will cool the gas and the effect will be an increase in density, the
cooling effect being much larger than density reductions due to the lower molecular weight of steam. The fan
curve will be changed in accordance with the fan laws, the denser gas resulting in the fan doing more work. In
fact this effect is so great that water injection will result in an increased mass flow of gas measured at a point
upstream of the injection point.
As an example, the original Hope Preheater fan curve has been taken. The system has been simplified to that in
Fig. 26.
Example
o
Initial fan gas flow = 4900 m3/min at 350 C
o
Specific heat of steam = 0.5 cal/g C
o
Initial water temp. 15 C
Problem
Given the above information what will the increase in kiln gas flowrate be on water injection ?
Step 1
o
Let new temp = T C
Heat balance:-
o
T = 230 C
31.6 623
× = 0.766 kg/m
3
New density = 0.64 x
32.7 503
Step 3 : Modify fan curve
Volume 0 1 2 3 4 5 6 7
Pressure 612.5 665 725 745 725 675 600 510
Power 290 410 525 630 720 790 840
0.766
Density ratio = = 1.195
0.64
Step 4
Here, we correct the new fan curve so that the volume it applies to are at the conditions upstream of the water
injection point. This is done because the system resistance line we draw in will refer to volumes at these
conditions.
o
The calculation is carried out by considering successive volumes of gas at 350 C reaching the injection point.
3 o
e.g. Volume flow = 3000 m /min at 350 C
o
Water required to cool.this to 230 C
1920
= 137 x = 83.8 kg/min
3140
= 887 mm wg
0 0 732
1000 872 787
2000 1744 846
3000 2616 887
4000 3488 885
5000 4360 850
6000 5232 735
7000 6104 705
A plot of the first column against the last column is then the requiread curve (curve 1K)
Assume P = kV2
(P = 680, V= 4900) is the original, operating point
Volume 4 5 6 7
Pressure 453 707 1019 1385
Then the new operating point will be at the intersection of the S.R.L. and the new fan curve.
i.e. 5400 m3/min
(intersection with curve 1K)
However, we now reach a contradiction in the calculation. if the kiln gas flow increased to 3460 kg/min,
then the addition of 137 kg/m, in of water would not quite cool it down to 230oC. Unfortunately, 230oC
is the temperature upon which we have based our fan curve calculations. To cool the extra kiln gas down
to 230oC we would need a further
320
137 x = 14 kg/min of water
3140
In other words we have assumed an injection rate of 137 kg/min but found that we in fact need 151
kg/min to cool the gas down to the temperature at which we have made our calculations.
Step 7
What we must do now is to make a better guess' at a water rate that we may assume to carry out our fan
curve calculation, so that the total water rate, calculated at the end of Step 6, 13 equal to the true water
rate.
137
New guess = 137x = 124 kg/min
151
We then use this value, 124, in place of 137 kg/min and repeat Steps 1-6.
Step 5 then gives us an operating point of 5350 m3/min (intersection with curve 2K) = 3420 kg/min.
280
=124 x = 11 kg/min
3140
This is sufficiently close to the true water rate (137) and further iteration is not necessary.
Power Drawn
When the operating point is at 5350 m3/min (curve 2K) the fan will handle 4700 m3/min (curve 2). At this
volume the fan shaft power will be 820 kW (curve 2P), whereas prior to water injection the fan power was 715
kW. Hence we may draw up the following table:-
Water injection, therefore, may in some cases be used as a means of uprating the useful Volume handled by a
fan.
If the extra gas that is drawn were not required one would, of course, damper back along the curve 2K, or reduce
the fan speed until the original kiln gas flow was attained.
Blue Circle Cement
Module 5
Section 2
4. Fan Curves
As a first step to an understanding of fan performance it is necessary to get a clear conception as to the
manner in which a fan actually operates.
The purpose of a fan is to establish and maintain between the fan inlet and fan discharge a difference in
pressure of the air or gas it is handling, and it is in consequence of this maintained difference in pressure
that, circumstances permitting, the volume of air or gas flows through the external circuit to which the
fan is connected.
The fan has to set up a difference in pressure sufficient to overcome the losses in the system, including its
own internal losses.
2.1 Energy
When considering fan problems it is seldom that the energies imparted to the air are referred to in terms
of potential and kinetic energy. Instead the term water-gauge is used (usually expressed in inches).
a) Potential Energy
The air within a duct system connected to a fan will be at a different pressure to the atmosphere
outside the duct and will have a capacity for doing work in virtue of that difference in pressure,
i.e.,'it will have potential energy.
b) Kinetic Energy
The air will be in motion and will have energy by virtue of that motion, i.e., it will have kinetic
energy.
c) Total Energy
The total energy is the algebraic sum of the potential and kinetic energies.
In all fan work the pressure of the atmosphere surrounding the fan at sea level is taken as the datum, and
pressures are considered positive or negative, according as they are above or below atmospheric pressure.
Therefore, since atmospheric pressure is taken as a datum and since the fan produces a difference in
pressure it follows that the air only possesses potential and kinetic energy by reason of being connected to
the fan, and therefore the energy must be supplied by the fan.
Within limits of the total energy in the air, the ratio of the amount of energy present as potential to the
amount of energy present as kinetic to the amount present as kinetic is governed entirely by the external
system to which the fan is connected.
This can be summarized as a fundamental principle:
“The resistance of the external system controls the volume which the fan can handle at a fixed speed."
2.2 Pressure
d) Static Pressure
Static pressure is the bursting or collapsing pressure according to whether the difference in pressure
between the inside and outside of the duct is positive or negative, that is, above or below atmosphere.
Thus the static pressure is a measure of the capacity of the air within the duct for doing work, and is
therefore a measure of the potential energy.
e) Velocity Pressure
Velocity pressure is a measure of the kinetic energy which the air possesses by virtue of its motion can be
measured by the Pitot Tube.
f) Total Pressure
Total pressure is defined as the algebraic sum of the static pressure and velocity pressure.
Since the velocity pressure can only be measured by impact it must always be a positive quantity, whether
the flow be on the discharge side of the fan, where the static pressure is positive, or on the suction side
where the static pressure is negative.
It follows, therefore, that upon the discharge side of the fan the total pressure will be numerically greater
than the static pressure, and upon the suction side of the fan the total pressure will be numerically less
than the static pressure.
g) Water-Gauge
The pressures to be measured are usually quite small and it is usual to refer the pressures to the height of a
column of water which could be sustained, hence the term "water-gauge".
In fan work, since the pressures generated are so small, the variations in air volume, due to compression or
expansion, are also quite small, and such variations are usually neglected. Variations in temperature and
barometric pressure, however, produce considerable -changes in volume and air density.
(A cubic foot of water weighs 62.4 lbs at 60 oF so that a cube of water with 12" sides would exert a
pressure of 62.4 lbs over the area of its base which is one square foot, so that a water-gauge of 12"
corresponds to a pressure of 62.4 lbs per square foot, or 1" water-gauge represents 5.2 lbs per square foot.
Hence to transform a pressure expressed in terms of inches water-gauge to a corresponding pressure in lbs
per square foot the water-gauge is multiplied by 5.2).
2.3 Manometers
The apparatus which actually indicates the water-gauge is known as a manometer (Fig. 2). For use on site
it generally consists of a plain glass U tube, one leg of which is connected to the duct where the pressure
is to be measured, the other leg being open to atmosphere. The pressure is then obtained by the difference
in the height of the water in the two legs of the tube.
The correctness of the readings of a manometer will be dependent upon the accuracy with which the
difference in level of the height of the columns of water in the two legs of the manometer is measured. It
is essential that the U tube be held in a vertical position. The difference in the levels of the water can then
be measured carefully by means of a scale attached to the gauge.
The water in the tube will tend to cling to the surface of the tube, and hence the surface of separation
between the air in the tube and the water will be cup-shaped, the smaller the bore of the tube the more
curved will be the surface of the water. If is usual to measure to the bottom of the cup-shaped surface (or
meniscus as it is called). For accurate work the cup-shaped surfaces must be the same, i.e. the tubes must
be of the same bore. It is usual to specify that the bore of such a U tube shall not be less than, say, 3/8”
diameter.
Fan total pressure PT, is the difference between the total pressures at the fan inlet and outlet.
Fan static pressure, PS, is the fan total pressure minus the fan velocity pressure.
Fan velocity pressure, PV, is the velocity pressure corresponding to the average velocity at the fan outlet.
i.e. the fan static pressure is NOT equal to the difference between the static pressures at the inlet and
outlet. It should be measured as the difference between the reading of a facing tube at the inlet and a side
tube at the far outlet (see Fig. 3).
Fig. 4 shows typical fan curves for each type. In general for a given impeller size, the pressure developed
increases with blade angle.
The blades are of short radial depth and consequently more of them (30 - 60 compared with 6 - 16 for
other fans) are required to have an effect on the air. It can be seen (Fig. 4) that a decrease in restriction
and hence an increase in airflow will cause an increase in power. The fan is said to be of the overloading
type and this effect may cause the motor to trip out.
These fans are unsuitable for dirty air, the narrowly spaced tips tending to bridge.
These have their main application in handling high dust burdens. They are of simple construction and
therefore the wearing parts can be of substantial material or even replaceable. The impeller is
unshrouded, but may have a back plate in the case of fibrous dust to prevent build-up on the shaft.
Disadvantages of the paddle fan are its low efficiency (about 60%), high speed of rotation (noisy
operation) and overloading power characteristics.
These fans combine some of the ruggedness of the paddle fan with the higher efficiency and higher
pressure development of the forward curve blade. The impeller is self-cleaning to a large extent and is
used extensively for dusty flue gases. Blades themselves are not easily made removable but replaceable
wear plates and nose pieces can be fitted if dust burdens are high.
Efficiencies higher than 75% are obtainable with this fan and it is therefore commonly used where high
dust burdens prevent the use of the more efficient aerofoil or backward inclined bladed fans.
3.3 Backward Bladed Fans
These fans are used mainly for clean air applications, although the use of a reinforced nose, or leading
edge, will allow low concentration of dust to be handled.
Efficiencies up to 90% are obtainable with fans having aerofoil section blades, and therefore the fan is
preferred for the larger machines. The high efficiencies are a result of laminar flow over the blade profile.
Backward bladed fans are non-overloading (see power curve on Fig. 4).
Summing up, the progressive use of fan types as the dust burden increases would be:-
Forward Curved
Backward Aerofoil
Backward Inclined
Forward Curved Radial Tip
Paddle
4. FAN CURVES
These show the effect on fan pressure, power and efficiency of a change in fan throughput (Fig. 5). The
horse-power on a fan curve is that which must be delivered to the shaft and does not take into account
losses in couplings and motors.
P may be either the fan static or total pressure and the efficiencies thus calculated are the static and total
efficiencies respectively.
The resistance of the system served by a fan depends upon the hardware in it and for any system there
will be a relationship between flowrate and pressure drop. If the system contains only dampers, ducting,
expansions etc., of fixed resistance, then the pressure drop may be taken as being approximately
proportional to the square of the flowrate (V2). This would not be true in the case of a resistance (as
opposed to pressure drop), that varied with flowrate, such as a cooler grate. The intersection of the system
resistance line with the fan pressure/ f lowrate curve gives the point of operation of the fan. The effect of,
for instance, closing a damper is to steepen the system resistance line, giving a new operating point (see
Fig. 6).
For a centrifugal fan there are three basic methods of controlling output-
a) Damper Control.
b) Variable Radial Inlet Vanes - these give a beneficial swirl to the air at the fan inlet and therefore
reduce the work the fan has to do.
c) Speed Control.
Initially, dampers are the cheapest to install and speed control the most expensive, but there is a reversal
in a running cost comparison. However, speed control usually only becomes economic on the larger
machines.
Whereas variation in a damper setting will change the system resistance and not the fan curve, the other
two methods produce new fan curves, that for speed control retaining the same fan efficiency.
Referring to Fig. 7, the effect of turndown to ¼ V and ½ V by the three methods is shown.
Graph A - Damper Control - Fan curve retained but system resistance line altered.
Graph B - Variable Inlet Vanes - New fan curves produced for each vane setting.
Graph D - Comparison of three methods on basis of power taken. (Note: loss due to coupling is
exaggerated.)
Graph C - Speed Control - New fan curves produced for each speed but efficiency retained.
This graph assumes no pressure drop across a fully open damper. This is a reasonable assumption for
some dampers operating in clean air, but would not be true for some ID fan dampers where build-up will
occur. The expected pressure drop across an open damper, based on experienced, should be taken into
account when deciding on the method of control.
Fan performance may be affected by worn impellers, inefficient drives, etc., but the most common causes
of fan performance being substandard are bends close to the inlet or outlet of the fan.
If the bend is at the inlet, the air is pushed over to one side of the eye, whereas a manufacturer's fan curve
is for uniform inlet conditions. Uniformity of flow at inlet is easily checked by a pitot traverse.
The best solution to this is an inlet box which can be sized so that the air enters the impeller uniformly,
failing that a square bend with cascade (Fig. 8) will give good performance. Retention of the original
bend and fitting of long turning vanes will improve matters greatly (Fig. 9).
Useful pressure in a fan originally comes from the acceleration of the air. As the air is again slowed down
by expansion, velocity pressure is converted into static pressure. For this to be done efficiently the
expansion should be slow and gentle and clearly a bend immediately on the fan outlet will not help
matters. This problem is not so easy to remedy but again a cascade will help.
APPENDIX I THE FAN LAWS
The performance of a fan may be changed by several methods, the main ones being a change in:-
a) Fluid density, p.
c) Fan size*. Most text books only quote the fan laws for “fans of geometric similarity", so that diameter,
width, outlet duct size, etc., are all increased at the same rate. Any convenient dimension may then be
taken as being representative of the fan, e.g. Impeller Diameter, Dim.
The effect of these on fan pressure, volume and horse-power is summarized in Table I.
Alteration In
Change In
Fan Pressure Volume Power
Density, p αp No change αp
Speed, n α n2 αn α n3
Fan Size*, Dim α Dim2 α Dim3 α Dim5
In order to help remember the above table, it is worth noting that fan pressure can be shown to be theoretically
proportional to the square of the tip speed whereas volume is proportional to the swept volume of the impeller
(fan size/volume is a cubic relationship since width increases as well as diameter). Power is proportional to the
product of volume and pressure.
The following is given as a guide to minimize losses in fan & associated plant efficiencies due to
mechanical/process deteriorations of fans.
1) Check to ensure that fan inlet & discharge ductwork is not restricted with material drop out ( or plates, etc.
left inside following major repairs.)
2) Check to ensure that control dampers are functioning correctly, indicating correct position externally ( eg.
with louvre type dampers that all louvres can move freely, and together
3) Check to ensure fan speed is correct ( eg. on belt drives check belts are not slipping. Check pulley sizes are
correct).
4) Check fan power absorbed at motor.
5) Check fan inlets to ensure.
a) At entrances no material deposits exist.
b) At fan eye that there exists a close concentric fit of duct to impeller, to eliminate recirculation, within the
fan casing.
c) Within casing, no wear exists at nose bridge causing turbulence.
6) Check to eliminate all inleaks at duct Joints, fan inlet joints, doors, etc., on fan suction side.
9) General inspections inside fan casing are necessary to ensure no obstructions exit, or wear of body, that
turbulence could be created.
The above checks are necessary at shutdowns, particularly at the beginning to assess work/repairs necessary
and also before start up to ensure correct repairs have been carried out, & no extraneous materials have been
left inside casings & ductwork.
In addition to the above physical condition checks, Technical Departments should carry out process fan
efficiency checks during normal stable fan operation. This is essentially to verify that the fans are operating on
their fan curves ( ie. as designed ) and therefore corrective measures can be planned/ programmed for
shutdowns if necessary.
The above data can then be compared with curve specifications using the fan laws.
Note coupling/ralley /belt efficencies compared to motor power mean that motor power is greater than fan shaft
( curve ) power.
If the operating point of the fan is found to be significantly off the design curve, it is suggested that the test is
repeated, for consistency, & than corrective actions/inspections can be programmed as necessary.
1.0 DEFINITION OF FAN PRESSURE
Fan total pressure, PT, is the difference between the total pressures at the fan inlet and outlet
Fan static pressure, PS, is the total pressure minus the fan velocity pressure.
u2
Fan velocity pressure, PV, is the velocity pressure corresponding to the average velocity at the fan outlet.
2g
i.e. the fan static pressure is NOT equal to the difference between the static pressures at the inlet and outlet. It
should be measured as the difference between the reading of a facing tube at the inlet and a side tube at the fan
outlet (see figure 1).
These show the effect on fan pressure, power and efficiency of a change in fan, throughput (Fig.2). The
horse-power on a fan curve is that which must be delivered to the shaft and does not take into account losses in
couplings and motors.
A calculation of the work actually done on the air will therefore give the efficiency of the fan itself.
5.2
Work done on air = pressure x flowrate x
33,000
PV
= HP (inches wg and cfm)
6360
P may be either the fan static or total pressure and the efficiencies thus calculated are the static and total
efficiencies respectively.
Because work is being done on the air, its internal energy will rise, and this, together with losses due to the
inefficiency of the fan will cause the air temperature to rise.
fan power
Temperature rise =
density × V × specific heat
PV × 5.2
× η
For air at ambient conditions =
779
η = total fan efficiency
(0.075 × V × 0.24)
P
= 0.37 oF
η
The performance of a fan may be changed by several methods, the main ones being a change in:-
a) fluid density, p
b) impeller speed, n (rpm)
c) fan size* most text books only quote the fan laws for “fans of geometric similarity”, so that diameter,
width, outlet duct size, etc., are all increased at the same rate. Any convenient dimension may then be
taken as being representative of the fan, e.g. Impeller Diameter, Dim.
The effect of these on fan pressure, volume and horse-power is summarized in Table I.
ALTERATION IN
CHANGE IN FAN VOLUME POWER
PRESSURE
Density, p αp No change αp
Speed, n αn2 αn α n3
Fan Size*, Dim α Dim2 α Dim3 α Dim5
Fan pressure may be either total, static or velocity. In order to help remember the above table, it is worth noting
that fan pressure can be shown to be theoretically proportiona1 to the square of the tip speed whereas volume is
proportional.
Module 5
Section 3
Jo/w I:: ‘I I~vm/~~w *d f.‘. Jrrb ‘liiil c-r, 7h Ah )il /l/“W,~ ccl.
1” -
I
Y
I 1 9 r I
I I
“Ii “ 2 01 0.7 0.1 01 .
I ” 0.3 0 3 0.4 U.3 0.2 0.1
$5 0, 05 0 5 0.J 03 ,I 7
,I ‘Jb,
/’
1: ---- .--
/ J
- .-
Blue Circle Cement
Module 5
Section 4
Fans
FL Smith
FANS
Abstract
In this lecture the most important fan types used in cement plants are presented and their advantages and
disadvantages are outlined.
The definition of fan efficiency and simple fan calculation methods are discussed and it is demonstrated
how to save energy by optimum fan regulation and modification of existing fans with high efficient
impellers.
Table of contents:
1. Introduction
5. Fan drives
5.1 Direct/indirect drive
5.2 Motor type and speed regulation
5.3 Drive efficiencies
6. Capacity regulation
6.1 Damper regulation
6.2 Inlet guide vanes
6.3 Fan speed regulation
7. Fan modifications
7.1 Impeller replacement
7.2 Energy savings
Fans
1. Introduction
Fans are used in all parts of a cement plant and the operational success of nearly all main machinery in the
cement plant is completely dependent on the reliability and capacity of the connected fans.
The specific electrical energy consumed of the total number of fan motors included in a modern cement
production unit normally amounts to 30-40 kWh/t clinker which makes the fan motors the biggest power
consumers at the cement plant after the mill motors.
Thus, it is very important that the fans included in a cement plant are very carefully selected and
dimensioned, and the plant operators must know how to operate the fans most efficiently.
Very broadly speaking the function of a fan is to create a flow of gas or air in the system to which the fan is
connected. In the fan the necessary energy to create the desired flow through the system is transferred to the
air or gas by a rotating impeller driven by the fan drive (f.inst. an electrical motor).
Fans are normally characterized as either axial or centrifugal fans. In an axial fan the flow is predominantly
axial, that is in the direction of the impeller axis whereas the flow in the centrifugal fan is predominantly
radial (that is perpendicular to the impeller axis) in the region where the energy transfer takes place.
In the cement industry centrifugal fans are the most commonly used and in the following we will therefore
only deal with centrifugal fans though much of the information, formulae etc. may also apply to axial fans.
A centrifugal fan can be categorized according to its various design details such as: inlet design, impeller
type, blade type, impeller position etc., see table 3.1.
Table 3.1: Design details which characterize a centrifugal fan.
Further, fans are often characterized according to their function or position in the system in which they
create a flow of gas or air, namely as:
2. Exhausters or induced draught (i.D.)fans, i.e. fans which remove air or exhaust gas
from a system
3. Circulation or recirculation fans are fans which maintain a closed loop air or gas flow
in a system.
In the cement industry the fans are often termed by the machine system to which they are connected, f.inst.:
In the following, four fan types commonly used as kiln I.D., and/or mill circulation fans in the cement
industry will be presented and the advantages and disadvantages of the individual types will be
explained.
Fig. 3.2 shows a centrifugal fan with open impeller. The air or gas is introduced in the fan in axial
direction, proceeds radially through the impeller and leaves the fan through a tangential outlet in the fan
casing (for the terminology used here,,-please refer to the legend of the figure).
The impeller has bolted on plane, radial blades which can be changed when worn or major changes in
the fan performance are required.
The impeller is overhanging, that is both shaft bearings are placed on the same side of the fan casing.
This provides for the use of a simple axial inlet to the fan, but the load on the bearing closest to the
impeller is very big and the bearing has to have a diameter corresponding to maximum shaft diameter.
Normally this fan type is designed for a fairly low specific speed, that is for a given rating the fan will
be designed with a relatively large impeller running with a low rotational speed.
The advantage of this type of fan is that with the radial blades the fear of dust accumulation on the
backside of the impeller blades is eliminated and the low rotational speed makes the impeller less
exposed to wear from dust contained in the gas being conducted through the fan. Further, should the
fan blades get worn, they are easily replaceable.
on the other hand this fan type exhibits a very low efficiency (65-70%) and the front side of the radial
blades are often exposed to hard coating formation when operating with kiln exhaust gas.
Due to its low efficiency this type of fan will probably only be used in special cases in the future, f-inst.
when the gas to be transported contains high amounts-of abrasive dust.
Fig. 3.2: Single suction fan with open overhanging paddletype impeller
Fig. 3.3 shows a centrifugal fan with closed overhanging impeller. The impeller blades extend between the two
impeller shroud plates also termed the cover plate and the back plate.
Thus, the gas or air flows through the impeller in "closed" channels.
This design allows for a wide variation of the blade form and inclination (angle with the impeller tangent).
Normally backwards inclined blades are used as this ensures a higher fan efficiency than radial and forward
curved blades.
The fan in fig. 3.3 has relatively short, moderately backwards inclined S-shaped blades. This blade form
increases the impeller stability, and at the same time reduces the impeller surface area exposed to dust
accumulation. The efficiency of this fan type is approx. 78%.
Fig. 3.3: Single suction fan with shrouded overhanging impeller with S-formed blades.
For a given tip velocity an impeller with backwards inclined blades exhibits at lower pressure rating than
impellers with radial or forward curved blades. Thus, a high pressure rating affords a very high impeller tip
speed for which it is difficult to design the fan shaft and bearings in case the impeller is overhanging.
Thus, for high pressure fans with backwards inclined blades the impeller is normally placed between the shaft
bearings.
Fig. 3.4 shows a high efficient fan designed according to this principle.
Fig. 3.4: Single suction fan with shrouded impeller with backwards curved blades suspended between the
bearings.
The fan shown in fig. 3.4 has backwards curved blades with a tip angle of 50o. The maximum efficiency of
this fan type exceeds 80% despite the inevitable pressure loss in the inlet box. It is designed for an impeller
tip speed up to 180 m/s.
The major draw back of this fan type is that when working with dust laden gas it is often exposed to soft
coating formations on the back side of the impeller blades and/or wear of the impeller due to its high
rotational speed. However, various solutions to remedy these draw backs are offered, such as pressurized air
cleaning systems, hard facing of the impeller blades etc.
For very large flow capacities double suction fans are used, see figure 3.5.
A double suction impeller can be considered as two (laterally reversed) single suction impellers placed on the
same shaft and having a common back plate. The impeller is placed in a fan casing having an inlet box on
each side. Thus, the double suction fan has two inlet ports, which makes the inlet ducting to such a fan a little
more complicated, but on the other hand the capacity of the double suction fan will correspond to two similar
single suction fans with the same impeller diameter, and thus the double suction fan will often be the most
economical solution for large capacities.
Some fan suppliers provide single suction as well as double suction fans with backwards inclined airfoil
blades. See figure 3.6.
With this blade form a very rigid rotor is obtained even with a small plate thickness in the blades. This blade
form also ensures a very high fan efficiency (up to 85%). on the other hand the thin blades will naturally be
very sensitive to wear.
Fig. 3-5: Double suction fan.(Deutsche Babcock)
In this section some basic definitions and "fan laws" are presented and methods for "manual" calculations of
fans are outlined.
Normally, fans are dimensioned using a number of computer programs available at the fan suppliers, and the
purpose with the discussion here is only to give an idea of the principles used for dimensioning of large fans.
Further the importance of distinguishing between static and total pressure rating and different fan efficiency
definitions is emphasized.
The operation of a fan (working with a clean gas) can be characterized by:
Q1' = the actual flow of gas to the inlet of the fan.
ϑ1' = the specific density of the gas of the inlet to the fan and
∆p t ' = the fan pressure rating expressed as the difference between the total pressure of the gas in the fan
outlet and the fan inlet, respectively:
∆p t = p t 2 − p t1
When using small p's for pressure we refer to the gauge pressure that is the pressure stated relatively to the
ambient pressure, B (barometric pressure), at the elevation at which the fan is placed.
P=p+B
The total pressure at a certain point in a flow of gas is the sum of static pressure, ps, and dynamic pressure, pd,
of the gas at the apoint considered
pt = ps + pd
pt is the pressure measured through the center hole in a pitot tube directed against the gas flow direction and p
is the pressure measured through the side holes of the pitot tube (perpendicular to the flow direction).
1
The dynamic pressure is calculated as pd = pt - ps = (ϑ)u 2
2
in which
This can be calculated from the total fan pressure rating ∆pt as follows:
1 1
∆p s = ∆p t − (ϑ 2 )(u 2 ) + (ϑ1 )(u 1 )
2 2
2 2
2 2
1 Q 1 Q
∆p t − (ϑ 2 ) 2 + (ϑ1 ) 1
2 A2 2 A1
Herein Q1 and Q2 designate the flow through the inlet/ outlet of the fan, respectively, and A1 and A2 the cross
sectional area of the inlet and outlet of the fan.
The actual gas density ϑ can be calculated, if the normal density ϑ N (that is the density at 0oC and 760 mm
Hg) of the gas is known, together with the temperature t and relative (static) pressure p:
273 B + p
ϑ = ϑN
273 + t 760
in which B (the barometric pressure) and p (with sign should be inserted in mm Hg (1 mmHg = 133.3 Pa).
Likewise an inlet flow expressed in normal cubicmeters QN can be transformed to actual flow volume Q by
ϑ 273 + t 760
Q = QN N = QN
ϑ 273 B + p
Assuming that the compression of the gas flow in the fan is an isentropic (or reversible adiabatic) process and
that the gas velocity is the same in the fan inlet and outlet we can calculate the fan power input (using a normal
thermodynamic nomenclature):
1 1 1 1
as ds = 0 (isentropic process)
The latter expression can for an ideal gas be integrated to (see ref.(l)):
k −1
k −1
k P2 k k ∆Pt
Pis = (m )(v1 ) (P1 ) − 1 = (Q1 ) (P1 ) 1 +
k
− 1
k − 1 P1 k − 1 P1
Here k is the adiabatic constant for the ideal gas k = c p / c v
1 ∆Pt k + 1 ∆P2
2
Pis = (Q1 )(∆Pt ) 1 − + ...... = ((Q1 )(∆Pt )(α ))
2k P1 6k 2 P1
P
The last factor is called the isotropic compressibility coefficient. The relation between α and t for atm.
P1
air (k = 1.4) is shown in figure below (from ref. (1)):
Fig. 4.2: Isotropic compressibility coefficient as a function of relative pressure rise.
P (Q )(∆Pt )(α )
ηis = is = 1
P P
in which P is the actual power delivered to the fan shaft.
∆P
As α is close to 1 when the relative pressure rise across the fan t is small it has been common practice
P1
to use the efficiency defined as
(Q )(∆Pt )
η1 = 1
P
for fans generally. This efficiency is also termed the inlet efficiency.
It must be noticed that η1 is not a real efficiency as η1 would be greater than 1.0 for an ideal (high pressure)
fan.
One should always be aware that the fan efficiencies stated by different sources might have a different
definition.
The dimensioning of large industrial fans is normally based on the experimental results from a performance
test with a smaller fan model similar to the large fan. The results from the model performance test might be
expressed as performance curves, as illustrated in figure 4.3.
The "left" (full line) curve establishes the relation between the total pressure increase ∆Ptm obtained across the
model fan as a function of the volume flow Q lm measured at the inlet to the fan.
Correspondingly the "right" (full line) curve establishes the dependence between the volume flow Q lm and the
fan inlet efficiency ηlm defined as
(Q )(∆Ptm )
ηlm = lm
Pm
in which p m is the measured fan power consumption (at the fan shaft) at the flow Q lm .
If the compressibility of gas flow through the fan can be neglected (that is assuming the volume of the gas
does not change when passing the fan) the performance of a fan similar to the model fan can be calculated by
means of the three basic relations:
3
n D
Q = (Q m ) (1)
nm D m
2 2
ϑ n D
∆Pt = (∆Ptm ) (2)
ϑm n m Dm
η = η m = ηlm (3)
Qm is the flow through the model fan (= Q ml for incompressible flow)
Q is the flow through the actual fan
∆Ptm is the total pressure rating of the model fan
∆Pt is the total pressure rating of the actual fan
D is the actual fan diameter
Dm is the model fan diameter
n is the rotational speed of the actual fan
nm is the rotational speed of the model fan
ϑ is the specific weight of the gas drawn through the actual fan
ϑm is the specific weight of the gas drawn through the model fan
η is the inlet efficiency of the actual fan and
ηm is the efficiency of the model fan
Thus, the flow/pressure curve at any rotational speed and gas density for any fan size in the same series as
the model fan can be calculated from the model fan curves using the "fan laws" (1), (2) and (3).
Likewise a flow/fan power characteristic curve can be calculated using the expression
∆P (Q )
P = t (4)
η
Figure 4.4 shows actual fan performance curves for a kiln I.D.fan, which has been calculated from the
model fan curves in figure 4.3. (Note the curve has been drawn by a computer which also corrects for the
gas compressibility, thus the curves derivates slightly from the curves found using the above mentioned
calculation methods).
The optimal fan (that is the fan operating with maximum efficiency) for given values of ∆Pt , Q and ϑ is
characterized by a diameter Dideal and a rotational speed n ideal
Dideal and n ideal can be calculated from (1) and (2) by insertion:
n D ideal
3
Q = ideal (Q mideal ) (5)
nm D m
ϑ n ideal
2
D ideal
2
∆Pt = (∆Ptmideal ) (6)
ϑm n m Dm
where Q mideal is the flow at which the model fan obtains maximum efficiency η m = η m max and ∆Ptmideal
is the corresponding model fan pressure.
Combining (3) and (4) we obtain the following expressions for direct calculation of Dideal and n ideal
ϑ ∆P
0.25
Q
0.5
D ideal = × tmideal (D m ) (7 )
ϑ m ∆Pt Q mideal
Q D m
3
n ideal = (n m ) (8)
Q mideal D ideal
once the ideal fan diameter has been calculated a standard fan diameter close to the ideal diameter can be
chosen. The corresponding rotational speed might be calculated by iteration using the model fan curves
and the expressions (1) and (2), as illustrated by the following example.
Example 4.1
A raw mill fan (working with dust free air) of the HAF type should be dimensioned with max. efficiency
at the following operational data:
Neglecting the flow compressibility in the model fan as well as in the actual fan the calculation goes as
follows:
From figure 4.3 we find that the model fan obtains max. efficiency at Q mideal = 6.5 m 3 / s corresponding to a
pressure rating of ∆Ptmideal = 4140 Pa.
100 1.1 3
n ideal = (1450 rpm ) = 932 rpm
6.5 3.17
Choosing the standard fan impeller HAF 315/315 and assuming a rotational speed of 940 rpm we find from (1)
n m D m 3 1450 1.1 3
Q m = (Q ) = (
100 m / s = 6.57 m / s
3 3
)
n D 940 3.15
ϑ n
2
D
2
0.83 940 2 3.15 2
∆Pt = (∆Ptm ) = (4120 ) = 9820 Pa
ϑ m n m Dm 1.2 1450 1.1
which is very close to the desired ∆Pt . Thus further iteration is not necessary and a HAF 315/315 operating at
940 rpm would be the right choice.
The inlet efficiency for the actual fan will be equal to the efficiency for the model fan at 6.57 m 3 / s which is
read to 81.2% thus the shaft power for the actual fan is calculated to
( )(
(Q )(∆Pt ) 100 m 3 / s 9810 N/m 2
P= 1
)
= = 1208 kW
η1 0.812
With a safety factor for the fan motor of 15% the fan motor must have a minimum shaft power of 1390 kW.
In practice a fan always works with compressible flow and a treatment using the above mentioned calculation
methods, which assume an incompressible medium, always introduces some error in the calculations.
As long as the fan works with a pressure difference ∆Pt which is approximately the same as the pressure
difference ∆Ptm read in the model fan curve used for the calculations, the error will be small.
However, for accurate dimensioning of high pressure fans it will be necessary to take the flow compressibility
into consideration.
Such calculations are quite complicated and they are normally always performed by means of computers.
Thus, we shall refrain from dealing with these here.
It can be mentioned, that recalculating example 4.1 considering the-gas compressibility results in the same
selection of fan diameter but the fan rotational speed is calculated at 926 rpm, and the fan power at 1170 kW
rather than the 940 rpm and 1208 kW, respectively, calculated on neglecting the gas compression.
Up till now we have assumed that the fans to be calculated operate with clean gas.
Many fans used in a cement factory operate with dust laden gases f. inst. all kiln I.D. fans and most mill air
circulation fans.
The dust concentration in the gas passing these fans are typically 30-60 g / m 3 .
Very little can be found in the literature concerning calculation of fans for gas with such a dust load.
If the dust is so fine that the dust particles could be assumed to pass the fan evenly distributed in the gas, the
fan could be calculated as a fan operating with (clean) gas with an equivalent (or apparent) gas density of
ϑleq = ϑ1 + s (9)
s is the dust concentration in the gas (in kg/m3)
In this case the dust loading of the gas would not affect the fan power input, but the calculations would
suggest the use of a slightly bigger fan operating at lower speed than in the case of the fan operating with (the
same volume of) clean gas.
However, the dust is known to concentrate along the front side of the impeller blades on its way through the
rotor and the dust particles - leaving the rotor with a velocity close to the tip speed of the rotor, will at least to
some extent hit the fan casing and thereby loose their kinetic energy (obtained in the rotor).
According to ref. (2) it will be a good approximation to account for the dust loading simply by multiplying the
fan power found for the fan operating with clean gas with the ratio between the apparent density of the dust
laden gas and the clean gas density thus
ϑ leq
Pdust = − 1(P ) (10 )
ϑ1
The following example will illustrate the order of magnitude of the "dust correction".
Example 4.2
If the fan from example 4. 1 was to operate with gas with a dust loading of s = 40 g / m 3 the approximate fan
power increase according to (10) is estimated to
0.83 + 0.04
Pdust = − 1(1208 kW ) = 58 kW
0.83
Thus according to this calculation the fan power increases with almost 5% 5rom 1208 kW to 1266 kW when
the gas is loaded with 40 g / m 3 dust.
5. Fan Drives
The fans used in the cement and similar industries are nearly always driven by an electrical motor.
The motor might either be connected directly to the fan shaft via a flexible coupling or the motor power might
be transmitted to the fan through a V-belt drive (up to 400 kW) or a gear unit.
If direct drive by A.C. motors shall be used the fan must be dimensioned for a synchroneous or asynchroneous
motor speed cf. table 5.1.
Table 5.1
In many cases a direct drive will not allow for dimensioning the fan with optimum efficiency at the actual
operational conditions and in this case an indirect drive must be chosen.
For large fans it will also some times be cheaper to choose an indirect drive via a reducing gear coupled to a
motor with high rotational speed (few poles) than to choose a direct drive with a slow running motor (many
poles).
To-day the following combinations of motor types, starters and speed regulation systems are used for (larger)
fans:
For smaller shaft outputs ( <100 kW) fixed speed squirrel cage motors with direct start will always be preferred
due to the low price and easy maintenance of the squirrel cage motor.
For larger motors the large starting current for the squirrel cage motor with direct start (4-8 times the full load
current) can be a problem.
The starting current can be reduced by using an Y/∆ starter, but also the starting torque of the motor is
reduced, which results in a larger starting time which might not be permissible. In this case a slip ring motor
with rotor rheostat starter has to be used.
Synchroneous AC motors are relatively seldom used and only for large outputs ( >-500 kW). One advantage
of this motor type is that it can act as (capacitive) phase compensator (regulating the magnitisation of the
rotor).
For variable speed drives slip ring motors with rotor rheostat regulator and DC motors with thyristor
converter are the drive types most often used.
The advantage of the slip ring motor with rotor regulator is the relative simplicity of the system, but
connected to a fan the regulation range is relatively narrow (down to 60-75% of full speed) and the
efficiency of the drive goes drastically down when the motor speed is reduced as the "slip power" is wasted
(as heat) in the rotor rheostat (see below).
The speed of the DC motor with thyristor converter (often termed thyristor drive) is controlled by regulating
the armature voltage delivered by the thyristor converter. The drive system is somewhat more complicated
than the slip ring motor with rotor rheostat regulation but the control range is normally very wide (down to
10% of full speed) and the drive system maintains a good efficiency even at low motor speed.
The drastical reduction of the efficiency of a slip ring motor drive at reduced speed can be avoided if the slip
power instead of being wasted in a simple rheostat is treated in a "slip recovery system" in which it is
rectified, subsequently converted back to AC with grid frequency and then fed back to the (stator) terminals
of the motor. Such a system is termed a subsynchroneous converter cascade (or just cascade regulation).
However, this system is only price competitive to the DC motor drive system for large motor output (above
1000 kW).
For motor outputs below 1000 kW a squirrel cage motor connected to a static frequency converter might be
chosen. This drive system combines the easy maintenance of the squirrel cage motor with the wide
regulation range and low regulation power loss of the DC drive system.
The electrical power used to drive a fan can be calculated if the necessary fan shaft power P is known by
means of
p
p el = 1
(η )(η
trans el )
η trans is the efficiency of the power transmission system
η trans 1.0 for direct drive and
η trans 0.98-0.99 for indirect drive (gear or V-belt)
ηel is the efficiency of the motor inclusive a possible speed regulation system.
For large fixed speed AC motors the efficiency el normally lies between 92% and 96%. For various variable
speed drives the efficiency as a function of the relative speed reduction ratio appears from figure 5.1.
Drive efficiency, %
The total pressure drop ∆Pt across most flow systems in-, creases approximately with the second power of
the flow through the system, that is
(
∆Pt = (K )(Q )
2
) (1)
At full speed the fan which transports the gas through the system is also characterized by an unique
relation between total pressure rating and gas flow
Thus, at full speed the fan will generate a certain flow, Q., through the system which can be found from
the equa,tion
((K )(Q ) ) = f (Q )
1
2
l (3)
or from the fan curve, cf. figure 6.1.
If it is desired to reduce the flow through the system this can be accomplished in a number of ways which
are more or less economical as seen from an operational point of view.
By aamper regulation the resistance coefficient in the system, K, is changed by opening or closing a damper
normally placed near the inlet to the fan.
Closing the damper somewhat will increase K and reduce the flow in the system according to (3) (assuming
that f(Q) is a decreasing function in the relevant flow interval)), cf. also figure 6.1.
In the example in figure 6.1 the flow through the fan can be reduced from Q1 to Q 2 by closing a damper in
the system so as to increase the total pressure drop in the system from ∆p t1 to ∆p t 2
By this regulation method the pressure rating of the fan increases, when the flow through the fan is reduced.
Therefore the power consumption of the fan only decreases slightly when the flow is reduced.
However, as the damper is a very simple and cheap flow control device this regulation method is normally
preferred for smaller fans with limited power consumption.
By means of inlet guide vanes placed just at the inlet to the fan the gas can be given a prerotation which - if
in the rotor rotation direction - reduces the pressure rating and the power consumption of the fan for a given
gas flow. Thus by changing the adjustment of the inlet guide vanes the fan curve (or the function f (Q)) can
be changed and thereby the flow through the fan can be reduced.
As the guide vanes not just introduce a pressure drop in the system, but also reduce the fan pressure rating
and power consumption for a given flow, the flow regulation by means of guide vanes is more economical
than simple damper regulation.
Fig. 6.2 illustrates this. The fan with inlet guide vanes operates in a system with the system curve K.
With fully open guide vanes the fan characteristic curve designated 0o applies. The resulting flow through
the fan is Q1 , and the power consumption P1 .
If the guide vanes are turned 45 o the flow through the system is reduced to Q 2 and the fan power
consumption to P2 .
If the flow had been reduced by a simple damper in the system, the considerably higher fan power
consumption P2 would have resulted.
However, guide vanes are more expensive than simple dampers and they are rather sensitive to wear when
operating with dust laden gas.
6.3 Fan speed regulation
Reducing the speed of the fan will reduce the pressure rating of the fan at a given flow through the fan, cf.
fig. 6.3.
If the fan is connected to a second order flow system in which the pressure drop varies with the flow
according to (1) the fan will maintain the same efficiency independent of the fan speed. That is, if the fan is
designed to operate with maximum efficiency when operating with full speed the fan will still operate with
maximum efficiency in the case the flow is reduced by changing the speed of the fan shaft.
Thus, this form of regulation gives a loss-free regulation of the fan itself, and the operating economy depends
on the losses in the speed regulation system.
Various methods can be applied for speed regulation of the fan shaft.
Smaller fans can be coupled to a fixed speed motor through a mechanical variator or a hydraulic or eddy
current coupling.
For larger fans motor speed regulation is nearly always applied to reduce the fan shaft speed.
If the motor is a slip ring motor with rheostat regulation the regulation loss is considerable whereas a near to
loss-free regulation is obtained with an AC motor with frequency convertors, slip ring motor with cascade
regulation and DC motors with thyristor rectifiers as explained in section 5.
In fig. 6.4 the fan power consumption as a function of the flow for the various flow regulation methods is
compared.
Fig. 6.3: Flow regulation by changing fan speed.
Fig. 6.4: Fan motor power consumption curves
Example 6.1
Q = 100 m 3 / s
∆p t = 9810 Pa
ϑl = 0.83 kg / m 3
It has been calculated that HAF 315/315 fan will provide the desired pressure and flow at a rotational speed of
n = 926 rpm. However, the fan is supplied with a directly coupled 1600 kW, 985 rpm slip ring motor.
The actual fan curves for the fan at 985 rpm and 926 rpm, respectively, are shown in fig. 6.5.
A (second order) system resistance curve corresponding to the above mentioned flow and pressure for which
the fan was dimensioned is also indicated on the diagram.
From the diagram it can be read that the fan operating at full speed, 985 rpm, and with fully open 3 damper
would produce a gas flow in the system of 106.7 m 3 / s and the total pressure difference across the fan would
correspond to approx. 11200 Pa.
The desired flow 100 m 3 / s can be obtained closing the fan damper somewhat to obtain the system resistance
curve K 2 shown in the diagram fig. 6.5.
From the diagram it can be read that the fan shaft power consumption hereby will decrease only slightly to
1395 kW. The pressure difference across the fan will increase to 11600 Pa of which (11600 - 9810) Pa = 1790
Pa is lost in the damper.
With a motor efficiency of 0.96 the electric power consumption of the fan regulated by damper is calculated at
1395 kW
Pel = = 1450 kW
0.96
If the motor speed instead is reduced to 926 rpm the desired flow would be reached with fully open fan
damper. From the diagram we read in this case a fan shaft power consumption of 1175 kW.
If the slip ring motor is rheostat regulated the actual motor efficiency (including losses in the slip ring
rheostat) can be calculated to
n 926
ηel = (η motor ) = (0.96 ) = 0.90
n nom 985
and the electrical power consumption of the fan would be
1175
Pel = kW = 1305 kW
0.90
If the motor was provided with cascade regulation (slip recovery system) the efficiency of the motor and
speed regulator would be 0.96 - 0.015 = 0.945. In this case the electrical power consumption of the fan would
be
1175
Pel = kW = 1245 kW
0.945
This example illustrates the importance of using speed regulation of the fan motor rather than damper
regulation even when a relatively small reduction of the gas flow in a system is desired.
Thus, in this case where the flow is reduced with only 6% we save 150 - 200 kW (depending on the variable
speed drive type) if the flow is reduced by reducing the motor speed rather than closing the fan damper
somewhat.
Fig. 6.5: Performance curve for Fan type: HA.F 315/315
(used for example 6.1)
7.0 Fan modifications
Rising electricity prices have created an interest in changing low efficient fans with high efficient fans.
However, to change a large existing fan with a complete new fan not only involves buying the new fan but due
to another design of the new fan the existing fan foundation cannot be used and it will normally also be
necessary to modify the ducting around the fan which increased the installation costs and the down time.
Another solution is to install a new impeller in the existing fan in order to increase its efficiency without
changing the fan casing, fan drives etc.
As an example we will describe the equipment which is now offered for modification of a fan with open radial
impeller like the one described in section 3.1.
Together with the new impeller an inlet cone and inlet nozzle are installed in the existing fan casing.
Further, it is necessary to modify the rear plate of the fan casing somewhat and install a new shaft sealing and
cooling disc (if the fan is to operate with hot gasses).
In many cases the existing shaft can be reused after a machining of the end on which the new impeller is to
be mounted.
As the new impeller has backwards curved blades, it will normally have to operate with slightly higher tip
velocity than the old impeller to obtain the required pressure rating.
Normally, this problem is solved by increasing the impeller diameter in relation to the existing impeller.
However, in some cases the casing does not allow for a sufficient increase in the diameter and in that case the
shaft speed has to be increased introducing a new gear box or by changing the gearing ratio of an existing
gear box.
As-the operational conditions for the fan which is to be modified are normally known very well, the new
impeller can be designed to operate with optimum efficiency under these conditions, and an efficiency of the
modified fan exceeding 82% can normally be guaranteed.
Comparing this with a typical efficiency of 68% of the existing fan with open impeller the relative power
saving by installing the new impeller can be calculated at
0.68
S = 1 − (100% ) = 17%
0.82
assuming the same drive efficiency before and after the modification.
If the existing fans always have been operated with reduced motor speed and/or with partially closed dampers
the new impeller can be designed to allow for operation with almost full motor speed and fully open damper.
In this case the relative power saving on installing the new impeller will be even greater.
Fig. : 7.2: Fan originally provided with open impeller after modification with new high-efficient impeller
etc. New parts indicated.
If on the other hand the existing fan is operating at full capacity and a further increase in the production
capacity of the system to which the fan is connected is desired, it will often be possible to design the
replacement impeller for a pressure rating which allows for the desired production capacity increase.
Finally, it should be mentioned, that if the existing open impeller is exposed to excessive wear, it cannot
be recommended to replace it with a high efficient impeller as this will probably be exposed to even
greater wear and does not allow for an easy replacement of worn impeller blades.
Example 7.1
Data for an existing raw mill fan with open ¿ 4500 mm impeller:
Budgetary quotation for new closed fan impeller etc. for modification of the fan:
0.67
1300 kW = 1040 kW
0.84
Module 5
Section 5
To J von Tress
From R.W. Davies, K E Rabson
Date April 16, 1995
Copies: Harleyville Works A Storey, K Middour, C Kessler
BCC Corporate K. E. Rabson
BCTC UK L P Evans, J Ediss
Subject: Harleyville Preheater Exhaust Fan Performance Verification
Please find enclosed a summary of the preheater exhaust fan performance measurements conducted in
conjunction with GEL, J Ediss and Harleyville personnel.
Refer to report reference JGE/JAS/USA/3 dated 20 March 1995 for electrical analysis performed by J Ediss.
Enclosed are
1. Exhaust Fan Curve calculations based on measurements and calculations by Rabson and Davies, calculated
fan inlet conditions are 198649 Acfm at 702oF and 23.6 inwg of suction. Calculated shaft power is
approximately 810hp which should be compared with electrical derived shaft power of 688hp.
2. Exhaust Fan Curve calculations based on measurements by GEL at 13h50 and calculations by Rabson and
Davies, calculated fan inlet conditions are 196123.6 Acfm at 702oF and 23.6 inwg of suction. Calculated shaft
power is approximately 803hp which should be compared with electrical derived shaft power of 682hp.
3. Exhaust Fan Curve calculations based on measurements by GEL at 191100 and calculations by Rabson and
Davies, calculated fan inlet conditions are 208424 Acfm at 702oF and 23.6 inwg of suction. Calculated shaft
power is approximately 954hp which should be compared with electrical derived shaft power of 792hp.
4. Back calculations from stack measurements indicate similar flows at the fan inlet of approximately 190000
Acfm.
As promised this information will be passed on to LP Evans and JG Ediss for their final evaluation.
Best Regards
c:\bccworks\hrI\phex1 804.sam
Blue Circle Cement
Module 5
Section 6
Module 5
PRESENTATIONS
• Bowmanville Calciner ID Fan Problem
• Fan Technology
• Centrifugal Fans and Systems
Blue Circle Cement
PRESENTATION
Density 0.61
Speed (rpm) 966
Diameter (in) 3.45
20 9.5 507.625
40 10.2 721.625
60 10.7 949
80 10.8 1149.63
100 10.5 1323.5
120 9.9 1443.88
140 9.0 1530.81
160 8.0 1604.38
180 6.7 1644.5
200 5.3 1671.25
220 3.8 1684.63
240 2.2 1698
2000
12.0
11.0
10.0
1500
9.0
8.0
Pressure kPa
7.0
Power kW
1000
6.0
5.0
4.0
500
3.0
2.0
1.0
0.0
0
0 2000 4000 6000 8000 10000 12000 14000 16000
Volume m3/min
1500
1 2 .0
1 1 .0
1 0 .0
9 .0
1000
8 .0
Pre s s u re k Pa
7 .0
Po w e r k W
6 .0
5 .0
4 .0
500
3 .0
2 .0
1 .0
0 .0
0
0 2000 4000 6000 8000 10000 12000 14000
Vo lu m e m 3/m in
Density 0.641
Speed (rpm) 849
Diameter (in) 3.45 8400.0 6.5
1500
1 2 .0
1 1 .0
1 0 .0
9 .0
1000
8 .0
Pre s s u re k Pa
7 .0
Po w e r k W
6 .0
5 .0
4 .0
500
3 .0
2 .0
1 .0
0 .0
0
0 2000 4000 6000 8000 10000 12000 14000
Vo lu m e m 3/m in
2000
12.0
11.0
10.0
1500
9.0
8.0
Pressure kPa
7.0
Power kW
1000
6.0
5.0
4.0
500
3.0
2.0
1.0
0.0
0
0 2000 4000 6000 8000 10000 12000 14000 16000
Volume m3/min
2500
14.0
13.0
12.0
2000
11.0 2121kW
9.6 kPa
10.0
9.0
Pressure kPa
1500
Power kW
8.0
7.0
6.0
1000
5.0
4.0
3.0
500
2.0
1.0
10080 m3/min
0.0
0
0 2000 4000 6000 8000 10000 12000 14000 16000 18000
Volume m3/min
PRESENTATION
Fan Technology
FAN TECHNOLOGY
COLIN PAXTON
FAN TECHNOLOGY
• FAN DESIGN
• PRESSURE DROP
• CONTROLLING AIRFLOW
Parts of a Fan
Casing
Inlet Duct Shaft
Impeller
Shaft seal
Drive Motor
bearing
Coupling
Inlet Eye
FANS
FANS
Impeller Designs
Power kW
Fan Operation
Volume and pressure
determined by fan and
resistance
Acceleration of gas
by impeller
Volume
Power
d1
P = k x (Velocity)2
Volume
Power
100,000
70,000 m3/hr
m3/hr
•Speed Control
Speed
Speed1000
750 rpm
rpm
Reduced Speed
Volume
Power Desired
Volume
Reduced Speed
PRESENTATION
• Centrifugal Fans
• Fan Laws and Fan Curves
• Fan Installation
• Fan Operation
• Fan Troubleshooting
Impeller
Shaft seal
Drive Motor
bearing
Coupling
Inlet Eye
Blue Circle Technical Training
Fan Operation
Operating volume is Volume and pressure
where fan and determined by fan and
resistance
resistance are
balanced
Acceleration of gas
by impeller
Power
20
300
15
200
10
5 100
0 0
0 10 20 30 40 50 60 70 80 90 100 110
Flow
2 3 (4-7)
Fan Speed, n ∝n ∝n ∝n ∝ log10 n ∝n
2 3 5
Fan Size, d ∝d ∝d ∝d
Total
Pressure Fan Static
Pressure
Total Pressure
Power
20
300 Static Pressure
15 Power
200
10
5 100
0 0
0 20 40 60 80 100
Flow
• Power increases
steeply over flow
range.
• SP curve shows area
of instability.
• Efficiency 80%.
• Power increases as
flow increases.
• SP curve is stable
across whole operating
range - important with
high turndown.
• Efficiency only 75%.
Other criteria more
important.
Blue Circle Technical Training
Axial Fan
• Combines SP curve of
BI fan with power
curve of FC fan.
• Efficiency 75%.
• Other criteria dictate
choice.
Predicting Performance
• Line of system
resistance intersects
fan curve. Intersection
is fan performance.
• Changes in system
gives different
performance.
• Fan selection
important.
Blue Circle Technical Training
Predicting Fan Performance
Away from Curve.
• Fan curves quoted at standard conditions.
• Fan relationships can be used to predict fan
performance at alternate conditions.
• Fan dimensions and speeds may also be
checked but consult manufacturer.
• Equations in notes.
• If air is spinning in
direction of fan, the
fan gives less lift.
• If spinning counter,
power and noise
increase.
• Difficult to evaluate.
• BI fans most
susceptible.
Blue Circle Technical Training
Improper Outlet Ductwork
• Does not allow for
efficient development of
static pressure.
• Best is 6 duct dias. of
straight duct.
• Elbows to be avoided,
best is in same direction
as fan.
• Expansion (evasé) often
used to increase static.
Fan Operation
• Best as efficiency is
maintained.
• Most expensive and
only justified on big
drives.
Fan Troubleshooting.
• Damper blades
twisted.
• Would exhibit high
pressure drop when
‘100%’ open.
• Build up on impeller
causes vibration and
(if excessive) high
power and low
efficiency.