See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/43277523

Calculated versus measured pressure losses for two seed cotton unloading
systems

Article  in  Applied Engineering in Agriculture · July 2001

DOI: 10.13031/2013.6463 · Source: OAI

CITATIONS READS
4 9,144

4 authors, including:

Greg Holt P.A. Funk

United States Department of Agriculture United States Department of Agriculture
328 PUBLICATIONS   2,227 CITATIONS    56 PUBLICATIONS   1,092 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Value Added Products from Agricultural Byproducts View project

Anti-inflammatory effects of α-MSH View project

All content following this page was uploaded by Greg Holt on 17 May 2014.

The user has requested enhancement of the downloaded file.

 CALCULATED VERSUS MEASURED PRESSURE LOSSES FOR
TWO SEED COTTON UNLOADING SYSTEMS
G. A. Holt, J. W. Laird, R. V. Baker, P. A. Funk

ABSTRACT.. Pneumatic conveying systems are commonly used in the cotton ginning industry. Inefficiencies in those systems
can reduce productivity, create choke ups, and result in high operating costs. The fact that inefficient pneumatic conveying
systems are costly is nothing new to the ginning industry. But how can system losses be determined and how accurate are the
equations recommended for calculating those losses? The purpose of this article is three–fold: 1) document modifications that
were made to the seed cotton unloading system at the USDA–ARS Lubbock, Texas ginning laboratory, 2) compare those
changes to values that were obtained from using standard friction loss calculations, and 3) report the effect that modifications
had on power consumption. The results showed that the equations ranged from 0.28 to 22.9% of the actual measured system
losses depending upon the calculation method used. Modifications resulted in a 19% velocity increase in the unloading
system’s suction telescope using 37% less power.
Keywords. Pneumatic conveying, Cotton, Friction loss, Cotton ginning.

A
critical aspect of any cotton gin is the seed cotton needed. Regardless of the seed cotton unloading system used,
unloading system. It is the unloading system that pneumatic conveying of seed cotton is utilized in the transfer
is responsible for transferring seed cotton from the of seed cotton into the gin for processing.
trailer or module into the gin for processing. An The handling of materials in a cotton gin is primarily per-
essential requirement of the unloading system is to convey formed by either centrifugal and/or axial fans. Both types of
the material into the gin at as constant and uniform a rate as fans represent one of the largest power consuming elements
possible (Laird et al., 1994). General components of a seed of a cotton gin, with pneumatic systems consuming 40 to 60%
cotton unloading system, depending on the type of cotton of the total power required to operate a cotton gin (Mangialar-
being handled, are: 1) a means of introducing seed cotton into di, 1977). The seed cotton handling system is one of the main
a suction conveying pipe, 2) vertical and horizontal energy consumers in a cotton gin (Watson et al., 1964; Antho-
conveying pipes, 3) a seed cotton separator, 4) a green–boll ny, 1989). Energy consumed from the handling of seed cotton
separator (optional), 5) an airline cleaner (optional), and 6) a will vary dramatically depending on factors ranging from
centrifugal suction fan or fans (Baker and Griffin, 1984). size and design of the gin, moisture and foreign matter con-
Currently, the two types of seed cotton unloading systems tent of the seed cotton, to operating and management proce-
predominately in use are pneumatic suction through a pipe dures (Baker and McCaskill, 1979). One of the factors
and module feeding systems. The pneumatic suction system affecting energy consumption of any pneumatic conveying
is one of the earliest mechanical means of bringing seed system is leaks. Some of the equipment utilized in cotton gins
cotton into the gin (Bennett, 1962). Over the years, today have a certain amount of inherent
leak" associated
automation has improved the early pneumatic unloading with their operation. For example, the green–boll traps, vacu-
systems into the swinging telescope suction system still in um droppers, air line cleaners, and suction separators are all
use today. The second type of unloading system is the cotton devices that will experience some air leakage. Typical leak-
module feeder. This system came about as a result of storing age rates have been established for most seed cotton process-
seed cotton in modules. With the advent of cotton modules, ing equipment. These standard leakage rates are based on
a means of bringing the seed cotton into the gin that addressed manufactures’ data or practical field experience. However,
problems specific to module storage and handling was excessive leaks over and above the standard rates can create
problems with proper conveying and handling of seed cotton,
resulting in less efficient energy use, and pose potential main-
tenance and operational problems resulting in increased
Article was submitted for review in October 2000; approved for downtime and reduced capacity.
publication by the Power & Machinery Division of ASAE in February
2001 .
In addition to air leaks, another factor affecting efficiency
The authors are Greg A. Holt, ASAE Member Engineer, Agricultural of an unloading system is pressure drop due to friction and dy-
Engineer, Joseph W. Laird, Agricultural Engineer, Roy V. Baker, ASAE namic losses in the ducts. These losses are important since the
Member Engineer, Research Leader, USDA–ARS, Cotton Harvesting and fan has to produce sufficient static pressure to overcome the
Ginning Research Laboratory, Lubbock, Texas, and Paul A. Funk, ASAE friction in the various system components. The total friction
Member Engineer, Agricultural Engineer, USDA–ARS, Southwestern
Cotton Ginning Research Laboratory, Mesilla Park, New Mexico. losses through any round duct are directly proportional to
Corresponding author: G. A. Holt, USDA–ARS, Rt. 3 Box 215, Lubbock, pipe length (the longer the pipe the greater the loss), inversely
TX 79403; phone: 806–746–5353; fax: 806–744–4402; e–mail: related to the diameter of the duct (the larger the pipe the
gholt@lbk.ars.usda.gov. smaller the loss), and proportional to the square of the veloc-

Applied Engineering in Agriculture

Vol. 17(4): 465–473 2001 American Society of Agricultural Engineers ISSN 0883–8542 465
ity of air moving through the duct (higher velocity results in For this study a friction factor of 0.02 was used since it is
squaring the loss). For dynamic losses, pressure drops are de- the value most commonly used, for galvanized sheet metal,
pendent upon the number and types of elbows as well as the in figures and tables referenced by cotton ginning industry
frequency with which the velocity of air changes as it flows personnel.
through the piping system (ACGIH, 1998; Murdock, 1996). In the case of Dynamic losses, two methods can be
Air flow friction losses in galvanized pipe are commonly de- employed: 1) the equivalent length of duct method or 2) the
termined by use of Air Friction Charts, which were loss coefficient method. For the equivalent length method,
constructed using the basic flow equation for pressure and the dynamic loss for the fitting (elbow, transition, branch,
loss in circular ducts (fig. 1). Values in the chart are based on etc.) is replaced with a length of duct that has an equivalent
standard air at a density of 1.203 kg/m3 (0.075 lb/ft 3) flowing loss (AMCA,1995). A friction chart (fig. 1) is then used to de-
through clean round galvanized ducts. Using the chart re- termine friction loss. Table 1, from the Cotton Ginners Hand-
quires knowledge of the duct diameter, and either the volu- book (Baker et al., 1994), shows an example of tables that can
metric flow rate or velocity of air in the duct. If the chart is be used in obtaining equivalent length values.
unavailable, the following Darcy–Weisbach Friction Coeffi- The loss coefficient method uses a dynamic loss coeffi-
cient equation could be used to compute the pressure losses cient, C, that relates the pressure loss in the fitting to the ve-
(Bleier, 1998): locity pressure at a given cross–sectional area of the fitting
(AMCA,1995). Equation 2 can be used to determine pressure
DP = K × (L/D) × Vp (1)
losses in fittings with known dynamic loss coefficients (Sev-
where erns and Fellows, 1958; Bleier 1998; Fan Engineering,
DP = pressure loss (Pa) 1983):
K = friction factor (dimensionless)
DP = C Vp (2)
L = length of pipe (m)
D = diameter of pipe (m) where
Vp = velocity pressure (Pa) DP = dynamic pressure loss (Pa)
C = dynamic loss constant (dimensionless)
Vp = velocity pressure (Pa)
The Dynamic loss constant is empirically determined for
components such as elbows, expansions, and transitions en-
countered in air duct systems. Figure 2 (ACGIH, 1998) shows
an example of dynamic loss constants for elbows of various
ratios of centerline radius and diameter.
This report describes the process and results of modifica-
tions made to improve the capacity of the unloading system
at the USDA–ARS cotton ginning laboratory in Lubbock,
Texas. The modifications were made in an attempt to in-
crease the velocity in the suction line while reducing the ener-
gy consumption. Modifications included replacing pipes
with larger, more appropriate sizes, installing smooth transi-
tion sections, replacing two fans in series with a single high
efficiency design fan, replacing an undersized Y–valve,
eliminating unnecessary excess elbows, and sealing leaks.
The results are compared with theoretical friction loss cal-
culations commonly found in the literature and used in the in-
dustry.

OBJECTIVES
To date, there have been few documented cases showing
the results of practical applications involving air line modifi-
cations in cotton gins, although proper sizing of pipe
Table 1. Straight pipe equivalent for 90³ elbows
of various throat radiuses.[a]
Throat radius of elbow in pipe
diameters 1/2 3/4 1 1–1/4 1–1/2 2 3
Length of straight pipe offering
Figure 1. Duct friction chart for round pipe in mm of water/m length of equivalent resistance in diameters 17 14 12 11 9.7 8.5 6.5
pipe (ACGIH, 1978. Reprinted with permission.) [a] For elbows less than 90_, the equivalent straight pipe resistance is
proportional to the bend (i.e. 30_ elbow will be one–third the value
shown).

466 APPLIED ENGINEERING IN AGRICULTURE

 the static and velocity pressures, the average air velocity was
calculated using the following equation:
V = Kp × Cp × (DP × (Ts/(Ps × Ms))1/2 (3)
where
V = velocity (m/s)
Kp = pitot tube constant (m/sec[(g/g–mole)
(mm Hg)/(°K)(mm H2O)]1/2)
Cp = pitot tube coefficient (dimensionless)
DP = average velocity pressure of air stream (mm H2O)
Ts = absolute duct temperature (°K)
Ps = absolute duct pressure of air stream (mm Hg)
Ms = molecular weight of air, wet basis (g/g–mole)
Calculations were corrected for pressure and temperature
Figure 2. Elbow loss coefficients for 90 degree round pipe. (ACGIH, 1998. using a pitot tube coefficient of 0.99 and a pitot tube constant
Reprinted with permission.)
of 34.97 (EPA, 1997). In conjunction with pressure readings,
measurements of rpm, power factor, volts, and amps were
to handle increased air volumes is widely recommended. performed on select days.
Even though air line modifications occur throughout the in- Calculating the pressure losses associated with the straight
dustry on a yearly basis, documentation has been limited for duct in the unloading systems was accomplished by using
the application of theoretical calculations to
real world" re- both equation 1 and figure 1. The dynamic losses (those asso-
sults. Therefore, the primary objective of this project was to ciated with elbows, transitions, fittings, etc.) were obtained
improve the efficiency of the telescope suction system at the from using equation 2 in conjunction with tables, similar to
USDA–ARS cotton gin lab in Lubbock, Texas while docu- figure 2, found in the literature referenced. To determine the
menting effects the changes had on pressure and energy con- dynamic losses associated with elbows, the equivalent length
sumption and compare the findings to friction loss and loss coefficient methods were used to calculate the fric-
calculations commonly used in the cotton ginning industry. tion losses. The first approach, referred to as the Equivalent
Length of Duct Method (ELDM), used the velocities at the
various locations and table 1 values in conjunction with fig-
EQUIPMENT AND PROCEDURES ure 1. For consistency, and due to the fact that velocity pres-
EQUIPMENT sure readings were not taken at all possible locations along
The equipment used to collect the data consisted of four the piping, the velocity or velocity pressures (Vp) needed to
aneroid gages, a standard pitot tube, tachometer, power factor calculate the dynamic losses were assumed to be constant for
meter, and a volt/amp meter. The aneroid gages had upper a given pipe diameter based on the first reading taken for that
limits of 127, 381, 635, and 1270 mm (5, 15, 25, and 50 in.) size pipe. For example, in table 2, the velocity used in calcu-
of water. A 457–mm (18–in.) long standard pitot tube was lating the dynamic loss for all the 406–mm (16–in.) diameter
used to measure the velocity pressure along with a commer- pipes in the Initial System Measurements and Calculations
cial tachometer for fan rpm. Power factor, volts, and amps section was calculated using the value of 2173.2 m/min
were measured using commercial power factor, volt, and amp (7130 fpm). The second approach, referred to as the Friction
meters. Loss Coefficient Method (FLCM), used equation 2 along
with the pressure loss coefficients, C, in figure 2.
MEASUREMENT AND CALCULATION PROCEDURES When calculating the dynamic losses for the expansions,
Procedures involved drilling holes into the duct work at two procedures were used. The first was a conservative ap-
selected locations and taking measurements while all equip- proach designed to overestimate the losses for each expan-
ment normally used in unloading seed cotton was in opera- sion–type transition. These transitions were evaluated as
tion. Ambient temperatures, relative humidities, and abrupt enlargements (or contractions) rather than gradual ex-
barometric pressures were obtained from the local National pansions. This approach was used in an effort to compensate
Weather Service (Lubbock Airport, two miles south of the gin for those suspected, but undetermined, air leaks in the system
lab) for each day and time measurements were taken. Veloc- not accounted for in the calculations. The second procedure
ity pressures were taken using either an 8– or 10–point tra- was to evaluate the transitions as per standard fluid flow liter-
verse across the cross sectional area of the duct. The use of ature. Dynamic loss constants for contractions and expan-
either the 8– or 10–point traverse was dependant on the sions were obtained from either tables or figures similar to
length of straight pipe located before and after the measure- figure 2 found in Jorgenson (1983).
ment location and the diameter of pipe being traversed. If Pressure losses associated with the various equipment
long straight runs of pipe existed, the 8–point traverse was used in the unloading system (i.e. rock and green–boll trap
used. If not, the 10–point traverse was used. Static pressures (R&GBT), air line cleaner, separator, etc.) were obtained
were also measured at several additional locations by means from actual measurements since losses in these devices can
of small holes drilled in the sides of the pipe. After recording vary dramatically depending upon circumstances ranging
from their operational settings, physical condition, etc.

Vol. 17(4): 465–473 467

 Table 2. Pressure drops, pipe velocities, and flow rates for the normal suction leg before and after changes.
Pipe Diameter Sp Velocity Flow rate Air Leakage
Location (mm) (mm H2O) (m/min) (m3/min) (% of final flow)
Initial System Measurements and Calculations
1.2 m after open end of suction 330.2 –88.9 1508.8 129.2 57.3
0.31 m before valve 20 355.6 –336.5 2230.5 222.1 26.6
0.31 m before air line cleaner Transition –349.3
0.31 m after air line cleaner 406.4 –393.7
Before transition into no. 2 separator 406.4 –419.1
0.31 m after no. 2 separator 406.4 –514.4
1’ above valve 27 406.4 –546.1 2173.2 282.9 6.5
At fan inlet transition NA –723.9
2.4 m after 2nd fan 508 146 1492.6 302.5 ––
After Change I
0.31 m above valve 27 406.4 –552.5 2232.6 289.6 6.2
0.31 m below valve 27 406.4 –660.4
At fan inlet transition Transition –698.5
2.4 m after 2nd fan 508 139.7 1523.7 308.8 ––
After Change II
1.2 m after open end of suction 330.2 –101.6 2037.6 176.2 48.1
0.31 m before valve 20 406.4 –317.5 1913.8 249.6 26.5
0.31m above valve 27 406.4 –596.9 2320.1 301 11.4
At fan inlet transition Transition –647.7
2.4 m after 2nd fan 508 190.5 1676.7 339.8 ––
After Change III
1.2 m after open end of suction 330.2 –76.2 1531.6 131.2 56.8
0.31 m before valve 20 406.4 –215.9 1611.2 209 31.2
0.31 m above valve 27 406.4 –419.1 1973 255.9 15.7
At fan inlet transition Transition –457.2
2.4 m after fan 508 165.1 1498.1 303.6 ––
After Change IV
1.2 m after open end of suction 330.2 –88.9 1753.8 150.2 50.4
0.31 m before valve 20 406.4 –190.5 1732.8 224.8 25.8
0.31m above valve 27 406.4 –419.1 2039.4 264.5 12.6
At fan inlet 508 –444.5
2.4 m after fan 508 147.3 1493.8 302.7 ––
After Change V
1.2 m after open end of suction 330.2 –88.9 1728.5 148 50.1
0.31 m before valve 20 406.4 –203.2 1670.3 216.7 26.9
0.31 m above valve 27 406.4 –431.8 2008.3 260.5 12.2
At fan inlet 508 –438.2
2.4 m after fan 508 139.7 1463.9 296.7 ––
After Change VI (Final)
1.2 m after open end of suction 330.2 –88.9 1879.4 160.9 45.9
0.31 m before valve 20 406.4 –215.9 1678.5 217.8 26.8
0.31m before air line cleaner Transition –279.4
0.31 m after air line cleaner 406.4 –304.8
Before transition into no. 2 seperator 406.4 –317.5
0.31 m after no. 2 seperator 406.4 –381
0.31 m above valve 27 406.4 –406.4 2009.9 260.7 12.4
At fan inlet 508 –438.2
2.4 m after fan 508 139.7 1468.2 297.6 ––

UNLOADING SYSTEM seed cotton through a free–air valve and R&GBT. From the
The USDA–ARS Lubbock, Texas ginning laboratory has R&GBT, seed cotton is conveyed in the normal suction leg
two unloading routes for seed cotton (fig. 3). Figure 3 illus- through three Y–valves, V1, V20, and V3, to an airline clean-
trates the two unloading system routes before any changes er and then to the suction separator over the automatic feed
occurred. The routes are referred to as the 1) normal suction control. The belt dryer leg is identical to the normal suction
leg and 2) belt dryer leg. Both routes start with the suction leg with the following exceptions: 1) from the R&GBT the
telescope unloading either a trailer or module and bringing seed cotton travels through Y– valves V1 and V2 to the sepa-
rator over the belt dryer, and 2) from the belt dryer

468 APPLIED ENGINEERING IN AGRICULTURE

 and 16–in.) diameter straight pipe, respectively. In addition
to the straight pipe, there were several elbows, transitions,
and Y–valves. The quantity, size, and degree of the other
components in the initial normal unloading system are listed
in table 4. Table 4 only includes those components used in the

every day" operation of the unloading leg (i.e. the airline
cleaner bypass is not listed). The components in the belt dryer
leg are not shown since they are similar to the normal unload-
ing system with only minor changes in dimensions.
The initial system used two centrifugal No. 45 fans in se-
ries to move the air. The fans were powered by 45– and
56–kW (60– and 75–hp) motors that operated at 1945 and
1975 rpm, drawing 54 and 75 amps, respectively. The initial
static pressure readings taken on both unloading legs along
with their corresponding velocity and volumetric flow rates
are shown in tables 2 and 5.

MODIFICATIONS
The changes made to the unloading system occurred over
Figure 3. Schematic of the two routes of seed cotton unloading system for a 13–month time period from July 1998 to August 1999. The
the USDA–ARS Lubbock, Texas, ginning laboratory. changes are referenced in the order they occurred from
Change I to Change VI. Locations of the various elements
the seed cotton drops into the feed control thus bypassing the changed can be visualized by referring to figure 3.
airline cleaner. The unloading system air from either separa- The first modification, Change I, involved replacing the
tor is conveyed through Y–valve 27 (V27), the unloading fan, 356–mm (14–in.) diameter pipe transition segments between
and to the unloading system cyclone. Table 3 shows the Y– the free–air valve, the R&GBT, and V1, with section of
valves in the unloading system and their respective functions.
The primary valves critical to the unloading system regard- Table 4. Unloading system components quantity, size,
less of the unloading leg used are V1, V3, and V27. The other and degree for nomal suction leg.
valves (V2 and V20) are used for research purposes and are Throat
not needed in the normal operation of the unloading system, Size Radius in
Component Quantity (mm) Degree Diameters
but are included in the discussion since leaks can occur as a
Free–air valve 1 406.4 × 355.6 –– ––
result of valves being in the system.
R&GBT 1 609.6 × 254 –– ––
Y–valves (V1, V3, 330.2 × 330.2
INITIAL SYSTEM LAYOUT V20 and V27) 4 openings –– ––
The initial system layout was comprised of 330–, 356–, 330.2 × 330.2 to
406–, and 508–mm (13–, 14–, 16–, and 20–in.) diameter gal- Transition (into ALC) 1 254 × 558.8 –– ––
vanized steel pipe. The length of duct used for each diameter 254 × 1270 to
pipe varied according to the unloading leg. The 330– and Transition (from ALC) 1 406.4 pipe –– ––
508–mm (13– and 20–in.) diameter pipe had 6.1 and 12.3 m Transition (into 406.4 pipe to
(20 and 40.5 ft) of straight pipe regardless of the unloading separator) 1 152.4 × 1803.4 –– ––
leg. For the normal suction leg, the length of 356– and Reducer from 508 round to
406–mm (14– and 16–in.) diameter straight pipe used separator blow box 1 406.4 pipe –– ––
throughout the system was 3.2 and 25.9 m (10.6 and 85 ft), Transitions (to and 330.2 × 330.2 to
from Y–valves) 3 355.6 pipe –– ––
respectively. Whereas the belt dryer leg consisted of 4.3 and
Transitions (to and 330.2 × 330.2 to
40.7 m (14.2 and 133.6 ft) of the 356– and 406–mm (14– from Y–valves) 2 406.4 pipe –– ––
406.4 pipe to
Table 3. Unloading system Y–valves and their function in the Increasors (to first fan) 1 482.6 round –– ––
USDA–ARS Lubbock, Texas gin lab.
Reducer (from fan 1 to 508 pipe to
Y–valves Unloading Leg Function fan 2) 1 482.6 round –– ––
V1 Both Allows selection of either the normal suction 381 × 457.2 to
or belt dryer leg. Transitions (from fans) 2 508 pipe –– ––
V2 Belt dryer Selects whether seed cotton goes to belt dryer Transitions 508 pipe to 1143
separator or into the transfer system for mov- (to cyclones) 1 × 285.7 –– ––
ing seed cotton from one trailer to another. Elbows 2 355.6 90 3/4
V3 Normal Allows the airline cleaner to be bypassed.
Elbows 4 406.4 90 1
V20 Normal This valve allows small lots of seed cotton in
Elbows 1 406.4 45 1
the gin to be picked up and introduced into the
ginning system without going out under the Elbows 1 406.4 60 1
suction shed or through the R&GBT. Elbows 4 406.4 30 1
V27 Both Works in conjunction with V1. Allows the pull Elbows 3 406.4 15 1
of the suction fan to be routed to the separator Elbows 3 508 90 1
of either unloading leg.

Vol. 17(4): 465–473 469

 Table 5. Pressure drops, pipe velocities, and flow rates for the belt dryer leg before and after changes.
Pipe Diameter Sp Velocity Flow Rate Air Leakage
Location (mm) (mm H20) (m/min) (m3/min) (% of final flow)
Initial Measurements and Calculations
1.2 m after open end of suction 330.2 –81.3
0.91 m before belt dryer separator 355.6 –241.3 1854.1 184.1 42.5
0.91 m after belt dryer separator 406.4 –292.1 1988.2 257.9 19.5
0.91 m above valve 27 (belt dryer leg) 406.4 –393.7 1863.5 241.7 24.5
At fan inlet transition Transition –723.9
2.4 m after 2nd fan 508 152.4 1580.4 320.3 ––
Final Measurements and Calculations
1.2 m after open end of suction 330.2 –76.2 1652.6 141.5 55.3
0.91 m after belt dryer separator 355.6 –317.5 2076.3 269.3 14.9
0.91 m above valve 27 (belt dryer leg) 406.4 –368.3 2121.7 275.2 13.1
At fan inlet transition Transition –400.1
2.4 m after fan 508 177.8 1562.1 316.6 ––

rectangular duct. Other changes included replacing the The last change, Change VI, entailed removing the
356–mm (14–in.) diameter pipe to V20 with 406–mm straight pipe insert in the fan inlet and sealing leaks in the sys-
(16–in.) diameter pipe and replacing the 356–mm (14–in.) tem with duct tape or caulk. Leaks were sealed in both un-
90° elbow after V1, in the belt dryer leg, with a 330–mm loading legs. The primary leaks sealed were the flanges on
(13–in.) flat back elbow. the rectangular ducts to and from the R&GBT, lids on the air-
Change II involved changes to the normal suction leg only. line cleaner and suction separator, and V1 and V20 flange
Modifications consisted of replacing all 406–mm (16–in.) connections.
pipe between V27 and the fans with 508–mm (20–in.) pipe
as well as replacing V27 with a larger Y–valve. The new Y–
valve had 406–mm (16–in.) square openings for the incom- RESULTS
ing lines from the two separators and one 457–mm2 (18–in.2)
For all intents and purposes, the belt dryer leg is the same
opening exiting into the 508–mm (20–in.) line. This installa-
as the normal unloading leg in regards to conveying the seed
tion of a new Y–valve resulted in enlarging the area in and out
cotton to the respective separators. Other than a few differ-
of V27 by 34 and 48%, respectively. The 406–mm (16–in.)
ences in transitions, the primary difference between the two
pipe between the fans was not replaced in this modification.
legs was in the use of the airline cleaner. In the belt dryer leg
These changes resulted in the elimination of eight 406–mm
the seed cotton does not pass through the airline cleaner,
(16–in.) elbows and increased the inlet and outlet area in V27
whereas in the normal suction leg it does. Therefore, the pri-
from 1090 to 1652 mm2 (169 to 256 in.2) and 2090 mm2 (324
mary discussions of results will focus on the normal unload-
in.2), respectively.
ing system with similar measurements having been obtained
The third modification, Change III, involved replacing the
for the belt dryer leg (table 5).
dual No. 45 fans with one No. 60 fan. The new fan was pow-
Table 2 shows the static pressure readings, calculated air
ered using the 56–kW (75–hp) motor that had previously
velocity and volumetric flow rate values, along with the per-
been used to drive the second dual fan. Since the new fan was
cent of leakage taken after each modification to the system.
larger, the belt drive ratio was changed in order to operate the
Using the initial system data in table 2, the friction losses, ac-
fan at the desired speed (1565 rpm). In addition to changing
cording to figure 1, associated with the straight runs of 330–,
drives, the existing V–belt drives were replaced with a cog–
356–, 406–, and 508–mm (13–, 14–, 16–, and 20–in.) diame-
belt drive to eliminate slip. After changing the fans, all pipes
ter pipe were 12.2, 13.0, 77.7, 14.5 mm (0.48, 0.51, 3.06, and
from V27 to the cyclones became 508 mm (20 in.) in diame-
0.57 in.) of water, respectively. Friction loss for the same pipe
ter. A total of four elbows and 21.6 m (71 ft) of straight pipe
as calculated by equation 1 were 12.4, 13.0, 84.8, and 16.5
between V27 and the fan completed the new installation. The
fourth elbow was installed at the fan entrance. Even though mm of water (0.49, 0.51, 3.34, and 0.65 in. H2O), respective-
ly. As expected, both methods produced almost identical re-
common practice states that either a straight run of pipe or a
sults. Differences between the values are more likely a result
banjo should be used at a fan inlet, we installed a 60³ elbow
at the inlet to determine its overall effect on the system. In of error in reading figure 1.
addition to replacing the fans and piping, the R&GBT was set
to operate with a 25.4–mm (1–in.) pressure drop. This adjust- INITIAL SYSTEM
ment to the R&GBT allowed proper operation while reduc- The initial measured pressure drop across the system,
ing excess air leakage. from the fan inlet to the suction telescope, was 635 mm
After changing the fan(s), the next two modifications in- (25 in.) of water. For the initial calculations, all transitions
volved slight adjustments to the pipe entering the No. 60 fan. were treated as abrupt enlargements or contractions. Using
Change IV consisted of installing 1.8 m (6 ft) of straight pipe equation 1 and the FLCM (eq. 2 and fig. 2), the breakdown
between the 60° elbow and fan entrance. The next change, V, of the calculated losses, for the initial system, in mm (in.) of
involved inserting straight pipe within half an inch of the fan water were 81.0 (3.19) for fittings and transitions,
wheel inside the fan inlet so as to force the air into the blast 183.6 (7.23) elbows, 126.7 (4.99) pipe, and 245.1 (9.65) for
wheel. the equipment. Where the equipment consisted of the pivot
for the suction telescope, free air valve, R&GBT, airline

470 APPLIED ENGINEERING IN AGRICULTURE

cleaner, and separator; the sum of all the calculated friction treating all of them as abrupt expansions and contractions) re-
losses utilizing the FLCM and the Darcy–Weisbach equation sulted in calculated friction losses of 371.1 mm (14.61 in.) of
was 636.8 mm (25.07 in.) of water, thus yielding a calculated water by the FLCM and 325.9 mm (12.83 in.) of water by the
estimate that was within 99.7% of the losses actually mea- EDLM, resulting in a 14.6% and a 0.63% over–estimation of
sured. When the losses were calculated using the ELDM the measured friction loss, respectively. Comparative results
(fig. 1 and table 1), a calculated loss of 576.8 mm(22.71 in.) for the calculated versus measured friction losses for the two
of water was obtained. This value was 9.2% less than the ac- unloading legs are shown in table 6. The increased variation
tual loss measured. The primary difference between the cal- of calculated versus measured losses for the belt drier can be
culated values was due to the sum of the elbow losses. The attributed to fewer measurements being taken for that leg
sum of elbow losses using the ELDM was 123.6 mm (4.87 in.) compared to those taken for the normal suction leg.
of water compared to 183.6 (7.23) obtained from the FLCM, On an individual component basis, when comparing the
a difference of 32.6%. actual pressure drop across the 406.4 mm (16 in.) 90³ elbows
[76.2 mm (3 in.) of water] with those calculated, FLCM
FINAL SYSTEM yielded the closest results (83% of actual) versus the method
Pressure loss across the system after all modifications and used in the Ginners’ Handbook (EDLM) which was only 58%
changes was 349.3 mm (13.75 in.) of water. Treating all tran- of actual. However, when comparing the measured loss for
sitions as abrupt expansions and contractions, the FLCM re- the three 508–mm (20–in.) diameter elbows with the calcu-
sulted in calculated losses in mm (in.) of water of 63.2 (2.49) lated values, the methodology outlined in the ELDM resulted
for fittings and transitions, 89.4 (3.52) for elbows, 84.3 (3.32) in a 26% overage whereas FLCM yielded a 61% overage.
for pipe, and 114.8 (4.52) for processing equipment for a total These results indicate that any specific method will be limit-
friction loss of 351.5 (13.84). Whereas the ELDM calcula- ed by assumption used in the calculations, measurement
tions yielded a total loss of 338.3 mm (13.32 in.) of water. locations, and the presence of flow disturbances before or af-
Thus the FLCM results were 0.65% higher than the actual ter the elbows. The variations illustrated for the elbows can
losses and the ELDM results were 3.1% less than the actual be attributed to all the factors listed. For example, the dynam-
losses. ic loss coefficients are measured with straight runs of pipe be-
Using the FLCM and the literature recommendations for fore the elbows. However, in this system there were at least
gradual expansions produced a calculated pressure loss of three instances where an elbow was attached to another el-
330.2 mm (13.0 in.) of water, after modifications, which is bow, which would have resulted in a change not accounted
within 94.5% of the actual loss. This result was opposite of for in figure 2.
what was expected. The expectation was that the final system
measurements would be more closely approximated by using Table 6. Comparison of actual versus measured friction losses for both
the literature recommendations for transitions and fittings as the normal and belt drier unloading system at the USDA–ARS
Lubbock, Texas ginning laboratory.
opposed to the conservative approach of treating them as
Measured
abrupt expansions and reductions in cross sectional area. The Calculated Loss
initial assumption was that the variation between the conser- Unloading Time Method of Loss (mm % Over/Under
vative approach and the literature recommendation for grad- System Period Calculation (mm H2O ) H2O) Measured Loss
ual expansions, before and after modifications, would be Initial FLCM 25.07 25.00 0.28
attributed to the leaks in the system thereby resulting in the Initial ELDM 22.71 25.00 –9.16
conservative approach yielding a greater deviation from the Final FLCM with
measured losses. The thought was, before any changes were abrupt
made, system leaks such as pipes not being joined properly, transitions 13.84 13.75 0.65
Normal Final ELDM with
worn out gores (seams) in the elbows, improperly operating
Suction abrupt
valves, etc., would result in leaks not accounted for in the cal- Leg transitions 13.32 13.75 –3.13
culations. Thus, the conservative approach would produce a Final FLCM with
closer estimation of pressure loss. Conversely, it was be- gradual
lieved that after modifications, the leaks would be repaired transitions 13.00 13.75 –5.45
or at least minimized to the point that the
recommended" Final ELDM with
gradual
equations would yield the closest estimation to the measured transitions 12.48 13.75 –9.24
system pressure loss. However, the results from the normal Initial FLCM 24.84 25.30 –1.82
unloading system indicated otherwise. The difference was a Initial ELDM 21.83 25.30 –13.72
combination of the non–quantified leaks and the assumptions Final FLCM with
used in the equations since the coefficients used are based on Belt abrupt
straight smooth runs before the elbows or transitions. Devi- Drier transitions 15.67 12.75 22.90
ations from the conditions for which the coefficients were es- Suction Final ELDM with
tablished will change the factors thus changing the calculated Leg abrupt
transitions 13.88 12.75 8.86
friction losses. Also, the number and location of measure- Final FLCM with
ments obtained has an impact on the results. gradual
It should be noted that results obtained from evaluating the transitions 14.61 12.75 14.59
belt drier leg where just the opposite from those obtained for Final ELDM with
the normal unloading system. The belt drier leg had a final gradual
transitions 12.83 12.75 0.63
system measured pressure drop of 323.9 mm (12.75 in.) of
water. Evaluating transitions as per the literature (i.e. not

Vol. 17(4): 465–473 471

ENERGY SAVINGS leak; and 5) the lids on the airline cleaner were not sealed as
Table 7 shows the power usage and fan operating data for they should be to optimize its use.
the unloading system before Change I, after Change II, and As a result of eliminating restrictions, excess elbows, and
after Change VI that was the final modification to the system. leaks we were able to increase the initial velocity in the tele-
The data emphasizes the benefit of eliminating restrictions, scope from 1508.8 m/min (4950 fpm) to a final velocity of
excess elbows, and leaks. Initially the system used 88.7 kW 1879.4 m/min (6166 fpm) while reducing power require-
(119 hp) to operate the two No. 45 fans in a series. The me- ments by 33 less kW (44.3 hp). Assuming a cost of 7 cents per
chanical efficiencies for the fans were below 50%. After kilowatt–hour, this improvement in system performance
Change II, the power requirements increased to 97.2 kW would result in an energy savings of $2.31 per hour of opera-
(130.4 hp) due to less restrictions in the line resulting in mov- tion.
ing more air. The mechanical efficiencies improved slightly
(less than 2.5%) as a result of the modification. After the final
modification, the reduction in system static pressure allowed SUMMARY
the suction system to be operated by one 55.9–kW (75–hp)
Restrictions, leaks, and excess elbows or fittings can
motor instead of the two motors that were used initially [44.7
greatly effect the performance of any pneumatic conveying
and 55.9 kW (60 and 75 hp)]. The power required was de-
system. Even though this fact is commonly understood and
creased to 55.8 kW (74.8 hp) without compromising the vol-
acknowledged throughout the cotton ginning industry, there
ume of air necessary to move the seed cotton. The mechanical
is limited amount of documentation illustrating the effect that
efficiency was improved approximately 5 percentage points
leaks, restrictions, and excess or improperly sized fittings can
from the initial efficiency.
have on system performance compared to values obtained us-
Throughout the process of improving the unloading sys-
ing standard calculations. This article illustrates how stan-
tem, the items that had the largest effect on pressure drop
dard pressure loss calculations, commonly used in the cotton
were: 1) V27 not properly closing thus resulting in a 76.2– to
ginning industry, compare to actual pressure losses encoun-
254–mm (3– to 10–in.) pressure drop across the valve de-
tered in the unloading system at the USDA–ARS Lubbock,
pending upon the suction leg being used; 2) the original suc-
Texas ginning laboratory. Comparisons between the actual
tion line was plumbed in such a way that it
snaked" around
and calculated system pressure drops were performed while
the other duct work going to other fans and equipment caus-
making various changes to the seed cotton unloading system.
ing an additional eight elbows to be used in the line between
Since the unloading system in the gin lab can either route the
V27 and the fan; 3) many of the elbows in the suction line
seed cotton through the
normal" suction leg or the belt dryer
were worn to the point of leaking at the gores (seams); 4) the
air gap in the R&GBT was set improperly causing excess leg, both of these legs were evaluated during this study.
There were two means of calculating pressure losses asso-
Table 7. Power and fan data before and after
ciated with the straight duct, elbows, and transitions. The
modifications to the unloading system. straight pipe friction losses were determined either by using
44.7 kW 55.9 kW the Darcy–Weisbach Friction Coefficient Equation or by the
(60 hp) (75 hp) use of a friction loss chart (fig. 1), which is derived from using
Motor Motor the Darcy–Weisbach equation. Losses due to elbows, transi-
Volt readings before any changes 464 464 tions, and fittings where calculated using both the Dynamic
Amp readings before any changes 54.00 74.67 Loss Coefficient Equation in conjunction with dynamic loss
Power factor measurements–initial 0.94 0.80 coefficient chart (fig. 2) or the Darcy–Weisbach Friction Co-
Fan rpm 1945 1975 efficient Equation in conjunction with equivalent diameter
Fan total pressure – initial (mm of water) 4419.6 5334 table (table 1). Pressure drops associated with processing
Power output (kW) 18.1 21.8 equipment were measured directly to help reduce variability
Motor output (kW) 40.79 48.01 that can arise due to varying operational settings, physical
Mechanical efficiency (%) 44.44 45.41 condition, and manufactures specifications.
Volt readings after change II 464 464 The overall measured static pressure drop from the inlet
Amp readings after change II 60.50 80.23 of the suction telescope to the fan, for both suction legs before
Power factor measurements–change II 0.94 0.80 and after changes, in mm (in.) of water were 635 (25.0) for
Fan rpm 1945 1975 the normal suction leg, 642.6 mm (25.3) for the belt dryer leg,
Fan total pressure – change II (mm of water) 4419.6 5334 349.3 (13.75) for the normal suction leg, and 323.9 (12.75)
Power output (kW) 20.41 24.63 for the belt dryer leg, respectively. This represents a system
Motor output (kW) 45.70 51.58 pressure reduction of 287 mm (11.3 in.) of water for the nor-
Mechanical efficiency (%) 44.66 47.75 mal suction leg and a 320 mm (12.6 in.) of water for the belt
Volt readings after all changes — 474 dryer leg. Therefore, modifications resulted in a 45% reduc-
Amp readings after all changes — 82.00 tion in pressure drop for the normal suction leg and a 49.6%
Power factor measurements–change VI — 0.83 reduction in the belt drier leg. Using standard equations to
Fan rpm — 1565 calculate the pressure loss, the equations generated pressure
Fan total pressure – final (mm of water) — 6934.2 losses that were within 0.28 to 22.9% of actual losses. Over-
Power output (kW) — 28.05 all, the equations produced values that would allow someone
Motor output (kW) — 55.88 evaluating a system to estimate losses fairly accurately de-
Mechanical efficiency (%) — 50.20
pending upon the assumptions used. It should be noted that
variations between the calculated values and those actually
measured can be attributed to a number of factors including:

472 APPLIED ENGINEERING IN AGRICULTURE

1) accuracy of the instrumentation used to obtain the mea- AMCA(Air Movement and Control Association, Inc.). 1998. In Air
surements; 2) variation in the layout of the duct work from Systems. Publication 200–95. Arlington Heights, Ill.
that assumed in the literature (i.e. elbows connected to el- Anthony, W. S. 1989. Reducing gin energy consumption and costs.
bows, etc.); 3) seed cotton cleaning machinery not properly Proc. Beltwide Cotton Conference 2: 543–545.
Baker, R. V., and A. C. Griffin, Jr. 1984. Ginning. In Cotton, eds. R.
operating; 4) leaks in elbows, duct work, and valves; and 5)
J. Kohel and C. F. Lewis, 401–405. Madison, Wisc.: American
fan performance and wear. All factors considered, pressure Society of Agronomy, Inc., Crop Science Society of America,
losses in a system can be calculated reliably. However, when Inc., and Soil Science Society of America, Inc.
optimizing a pneumatic conveying system in a cotton gin, ex- Baker, R. V., E. P. Columbus, R. C. Eckley, and B. J. Stanley. 1994.
perience and knowledge of the seed cotton cleaning equip- Pneumatic and mechanical handling systems. In Cotton Ginners
ment are vital in obtaining the most effective system Handbook. Agricultural Handbook No. 503. Washington, D.C.:
performance. USDA.
The effects of leaks and system inefficiencies can be very Baker, R. V., and O. L. McCaskill. 1979. Energy conservation in
costly. A simple routine evaluation of any pneumatic system cotton ginning. ACS Symposium Series, No. 107.
in a gin would be prudent. Potential problem areas include: Bennett, A. B., 1962. Cotton ginning systems and auxiliary
developments. Dallas, Tex.: Texas Cotton Ginners’ Association
1) undersized fittings and pipe (most commonly brought
and Cotton Gin and Oil Mill Press.
about by increasing the fan speed and air flow without being Bleier, F. P. 1998. Fan Handbook: Selection, Application, and
aware of how it affects the rest of the system); 2) worn and Design. New York: McGraw–Hill.
malfunctioning valves; 3) plumbing (have the runs as straight Environmental Protection Agency (EPA). 1997. Code of federal
as possible); and 4) leaks in elbows, pipe, and seed cotton regulations: Title 40, Chapter 1, Part 60, Appendix A, 579–581.
processing equipment. Optimizing any materials conveying Washington, D.C.: U.S. Govt. Printing Office.
system can only result in a more cost effective operation. Jorgenson, R. 1983. Fan Engineering. Buffalo, N.Y.: Buffalo Forge
Conventional engineering techniques commonly used to cal- Co.
culate static pressure losses, along with experience and Laird, J. W., B. M. Norman, S. Stuller, and P. Bodovsky. 1994. Seed
knowledge of the system, are effective means of accurately cotton unloading system. In Cotton Ginners Handbook.
Agricultural Handbook No. 503. Washington, D.C.: USDA.
estimating system losses and performance. As a result of the
Mangialardi, G. J., 1977. Conserving energy. In The Cotton
modifications made in this study, we were able to increase the Ginners’ Journal & Yearbook. 45(1): 13–16.
velocity in the suction telescope 19.7% using 37% less Murdock, J. W. 1996. Mechanics of fluids. In Marks’ Standard
power. Handbook For Mechanical Engineers, 10th Ed., eds. E. A.
Avallone and T. Baumeister III. 47–53. New York:
McGraw–Hill.
Severns, W. H., and J. R. Fellows. 1958. Air Conditioning and
REFERENCES Refrigeration. New York: John Wiley & Sons, Inc.
American Conference of Governmental Industrial Hygienists Watson H., A. C. Griffin, and S. H. Holder. 1964. Power
(ACGIH). 1978. Industrial Ventilation: A Manual of requirements for high–capacity cotton gins in the
Recommended Practice, 15th Ed. Lansing, Mich.: Committee on Yazoo–Mississippi delta. ARS 42–94. Washington, D.C.:
Industrial Ventilation. USDA.
_____. 1998. Industrial Ventilation, 23rd Ed.. Lansing, Mich.:
Committee on Industrial Ventilation.

Vol. 17(4): 465–473 473

 474 APPLIED ENGINEERING IN AGRICULTURE

View publication stats