Pneumatic Conveying Assignment (Final)

ASSIGNMENT/ASSESSMENT ITEM COVER SHEET Student Name:

FIRST NAME FAMILY / LAST NAME

Student Number: Email:

Course Code Course Title

Campus of Study:

Assessment Item Title: Due Date/Time: Tutorial Group (If applicable): Word Count (If applicable):

Lecturer/Tutor Name:

Extension Granted: Yes No Granted Until: Please attach a copy of your extension approval NB: STUDENTS MAY EXPECT THAT THIS ASSIGNMENT WILL BE RETURNED WITHIN 3 WEEKS OF THE DUE DATE OF

SUBMISSION

I declare that this assessment item is my own work unless otherwise acknowledged and is in accordance with the University’s academic integrity policy available from the Policy Library on the web at http://www.newcastle.edu.au/policylibrary/000608.html I certify that this assessment item has not been submitted previously for academic credit in this or any other course. I certify that I have not given a copy or have shown a copy of this assessment item to another student enrolled in the course. I acknowledge that the assessor of this assignment may, for the purpose of assessing this assignment: • Reproduce this assessment item and provide a copy to another member of the Faculty; and/or • Communicate a copy of this assessment item to a plagiarism checking service (which may then retain a copy of the item on its database for the purpose of future plagiarism checking). • Submit the assessment item to other forms of plagiarism checking. I certify that any electronic version of this assessment item that I have submitted or will submit is identical to this paper version. Turnitin ID: (if applicable)

Signature: _____________________________________________________________ Date: ___________________

Please tick box if applicable

Students within the Faculty of Business and Law, Faculty of Science and Information Technology, Faculty of Engineering and Built Environment and the School of Nursing and Midwifery: I verify that I have completed the online Academic Integrity Module and adhered to its principles

Students within the School of Education: "I understand that a minimum standard of correct referencing and academic literacy is required to pass all written assignments in the School of Education; and I have read and understood the School of Education Course Outline Policy Supplement, which includes important information related to assessment policies and procedures.

DATE STAMP HERE

Insert this way

http://www.newcastle.edu.au/policylibrary/000608.html

P n e u m a t i c C o n v e y i n g D e s i g n 3 1 4 6 5 6 8 P a g e | 1

The University of Newcastle

Due: 27/8/13

Pneumatic Conveying

Design MECH4220 - Bulk Materials Handling &Transport Joel Forbes 3146568


Abstract This report is concerned with the design of a system to meet a set of specified requirements

for the Pneumatic conveying of a bulk material. An important facet of design in these systems

is the need to minimise cost and maintenance of the system while meeting the required

system output. As such, the system will be designed around not only performance, but also

with the continual maintenance of the system in mind. This report aims to make final

specifications for the design of the system that best suits the above criteria.

The report makes extensive use of Microsoft Excel for the generation and iteration of the

calculations associated with pneumatic conveying systems. A set of data collected for a test

system was provided on which to base the building of a design model. From the test data,

empirical connections between system components could be generalised for similar ‘scaled

up’ versions of the test system. A method to forward calculate the solids friction factor based

on Froudes number and solids loading ratio was derived, as was an approximation of a

velocity lower limit to avoid blockage. From here, a design spread sheet could be generated

using these relations. The system was intended to:

Convey cement meal

Convey over 750 m

Have 2 long radius bends with a bend factor 0.5

Use air as the conveying fluid

The following design specifications were made to meet the requirements of the system

An overall pressure drop of 368kPa (3.68 bar)

A power consumption of 553kW

A steady state flow rate of cement meal of 27 tonnes/hr

A flow rate of air or 1.9kg/s

m* = 3.947

Positive displacement system for conveying

Dilute phase conveying

Schedule 40 pipes used

Pipeline stepped twice. First at 250m along the horizontal section from an 8 inch

nominal bore to 10 inch, then once more another 250m along the system (at the base

of the vertical rise) from 10 to 14 inch nominal bore.

Twin 4 m3 top discharge blow tanks in series (brand recommendation made)

Continuous conveying from the twin blow tanks.

Fluidising membranes fitted to the blow tanks

Use of a separate line and valve in depressurising the blow tanks

Use of supplementary air from air mover to fluidise blow tank material.

Oil free screw compressor (brand recommendation made) as an air mover

A removable cloth filter fitted on the receiving bin for gas-solid separation

An air pulse jet set up to reverse cleanse the filter

Potential addition of desiccant bed at air inlet for water vapour removal

Pressure gauge and velocity gauge at key parts of the system


Contents Abstract ...................................................................................................................................... 2

Introduction ................................................................................................................................ 4

Steps for Analysis of Test Data ................................................................................................. 5

The Testing System ................................................................................................................ 5

Model Derivation ................................................................................................................... 5

Model Checking ..................................................................................................................... 6

Results of Test Data ................................................................................................................... 7

Air Alone Friction Factor ....................................................................................................... 7

Particle Friction Factor ........................................................................................................... 8

Pick up velocity and solids loading ratio. .............................................................................. 9

Designing Process .................................................................................................................... 10

System Specifications .......................................................................................................... 10

Assumptions ......................................................................................................................... 10

Method ................................................................................................................................. 11

Final Design Parameters .......................................................................................................... 13

Comparison of Stepped and non-stepped systems ............................................................... 13

System Characteristics ......................................................................................................... 14

Recommendations .................................................................................................................... 17

General ................................................................................................................................. 17

Air Mover ............................................................................................................................. 17

Feeder ................................................................................................................................... 18

Gas-Solid Separation ............................................................................................................ 20

Maintenance ......................................................................................................................... 21

Conclusion ............................................................................................................................... 23

Appendix A – Equations .......................................................................................................... 25

Appendix B – Test Data ........................................................................................................... 26

Appendix C – Forward Calculation ......................................................................................... 27

Appendix D – System Parameters Table ................................................................................. 28

Appendix E – Uniform Pipe Bore Design Spreadsheet ........................................................... 28

Appendix F – Single Step Design Spreadsheet........................................................................ 29

Appendix G – Two Step Design spreadsheet .......................................................................... 30


Introduction Pneumatic conveying of material has widespread applications for the movement of material

in industry. The systems are on a based on simple fundamental operating principles and are

very suitable for the transport of powdered or granulated materials. The basic components of

a pneumatic conveying system include an air mover to supply the conveying fluid, a feeder to

introduce the solids to be conveyed, a pipeline, a receiving bin and a separation device to

recover the solids from the fluid. (Jones, Mills, & Agarwal, 2005)

The large scale of pneumatic conveyers means that cost, wear and maintenance are important

aspects to be considered in their design. Pneumatic conveyors in dilute phase are best

operated between 10 and 30 m/s of flow (Mills, 2004), but even within this optimal zone,

wear from the moving particle on pipes and system components is a major source of

maintenance requirement and spending. The pressure through the line is relatedly inductive

of high wear, so efforts at the design stage must be made to minimise these parameters. This

in turn minimises wear and its associated maintenance and replacement costs.

An important characteristic of pneumatic conveying systems used for design in this report is

that empirically derived relations between system parameters can be used for similar systems.

This effectively means that provided data for a smaller testing system can be used to derive

forward predictors of characteristics in the design of the final system. This report aims to use

this process to derive a design model, and from there select a system that meets requirements

while minimising cost, wear and maintenance.


Steps for Analysis of Test Data The approach taken in designing the conveying system first required data for a smaller testing

system to be analysed. From this, a model could be built that would be predict the behaviour

of the designed full scale system. Once the model had been derived and checked,

specifications of the full scale system could be fed into the model and key characteristics

altered to optimise performance.

The Testing System - 16 samples of data*

- Cement meal

- 176m long

- 14 long radius bends

- Bend factor 0.5

- 53mm inner diameter

- Fed by a 1 cubic metre blow tank

- Measured pressure drops were provided for varied flow rates of air and cement meal.

*The 16th

data sample was later discarded because of its incongruity with the rest of the data,

inferring it was either a measurement error or transient reading.

Model Derivation

The following is a stepped process of calculating system variables. All equations can be

found in Appendix A, along with the meaning of all variables used.

1. Calculation of the three Pressure values. Pressure at the outlet is known

(atmospheric), and a pressure drop is given. Therefore can be

solved to get Pin. Pavg is calculated using

2. Densities can be calculated from their corresponding pressure values (i.e. average

density uses average pressure)

3. In the same way, velocities can be calculated from their corresponding density values

by

4. Froudes and Reynolds numbers can be evaluated by Euations 13 and 6 respectively.

5. The Streeter equation (Streeter & Wylie, 1983) is then used to approximate particle

friction factor

=

6. The pressure drops in the system from fluid and bends can then be calculated using

the values from the previous steps, defined by equations 9 and 10 respectively.

7. Knowing that the given pressure drop is the sum of all pressure drops it can be said

that . The unknown pressure drop from the

solids can then be calculated.

8. Noting =

, the equation can be rearranged and solved for the

solids friction factor:


9. Hence evaluate S*= . This has no physical significance, but becomes useful

because it can easily be plotted against Fr (relevant below)

10. It is know that a relationship between and Fr exists, of the form

Therefore, the S* column calculated can be plotted against Fr and then curve fit to

find constants ‘b’ and ‘c’

11. Using these constants, generate the formula

√

which will be the forward calculator for the design spread sheet.

Model Checking When fitting the curve to obtain the solids loading factor, the R

2 value of the trend line should

be examined, as it is an indicator of the closeness of the curve’s fit, and thus, its accuracy as a

model. Acceptable curve accuracy should be 0.9; this will ensure that the error

associated with using the trend line for different variables will be acceptably small.

Aside from the accuracy of the trend line itself, the accuracy of the model’s forward

calculations needs to be investigated. To do this, a spread sheet (provided in Appendix C)

modelling the same system is set up identical to the first, but with two crucial differences:

- λs is evaluated using Equation 16 and the derived constants, not from its formal

definition in Equation 11

- No pressure drop is given. Therefore, one has to make an initial guess, run the

calculation, and then use the result as the new guess for iteration. This was done

automatically in Excel by setting the ‘guess’ cell to the result cell and letting the

program iterate.

The pressure drops that the system predicts using iterative calculation of Equation 16 can be

designated ∆Pcalc. The accuracy of this calculation can then be analysed by putting ∆Pcalc in

ratio with the pressure drops measured in the data (∆Pdata). This ratio should, if the

correlation is good, be close to unity (i.e. the provided and predicted data should be the

same). An acceptable level of error is approximately 20%, which should indicate that the

system, if designed fairly conservatively, will be sound.

1 0.2

By plotting the ‘r’ value for the test data against Fr and , the accuracy of the calculation

can be observed.

Similarly, a relationship between solids loading ratio and pick up velocity can be derived

from the test data, to act as an approximate test for blockage in the design system. On a

graph, m* and Ci can be plot against one another, and a curve drawn as a best fit for the data.

As none of the test data resulted in blockage, it is reasonable to assume that the rough curve

they form is a safe region. Of course, it is an approximation, and the actual pickup velocity

that causes blockage could be much lower, but it provides a rough guide to avoid blockage.


Results of Test Data As outlined by the method, characteristics of the pipe flow were determined using an excel

spread sheet (supplied in Appendix B).

Air Alone Friction Factor Using Equation 7, the following table was able to be generated from the model suggested by

Streeter (Streeter & Wylie, 1983):

=

Figure 1: Air friction factor for test data

This is a small portion of the total spread sheet generated in Appendix B, and shows the

differing friction coefficient for the tests of variant air and solid mass flow rates

Test Ma

(kg/s) Ms

(kg/s) λf

1 0.081 2.120 0.02170 2 0.081 2.380 0.02170 3 0.050 1.480 0.02288 4 0.047 1.560 0.02306 5 0.042 1.490 0.02341 6 0.038 1.530 0.02373 7 0.050 1.550 0.02288 8 0.034 1.550 0.02412

9 0.030 1.500 0.02459 10 0.026 1.300 0.02516 11 0.050 1.580 0.02288 12 0.022 1.100 0.02589 13 0.049 1.860 0.02294 14 0.032 1.750 0.02434 15 0.019 0.890 0.02658


Particle Friction Factor

Once the air alone friction factors were calculated, calculations of were able to proceed as

outlined by the method in the previous section.

Generating a graph of √ vs Fr number, a curve was fit of the appropriate form. The R2

value was very close to unity so can be considered accurate. Figure 2 shows the relation

√ or, in a more useful form

√

Figure 1: Empirical Determination of λs relation

With the new approximation of solids friction factor, a calculation of pressure drop could be

carried out for the same system, and compared to the actual measured pressure to analyse

accuracy.

Figure 2: Accuracy of the solids friction factor forward calculation: the ratio of calculated and measured pressure

drop plotted against solids loading ratio and Fr. The red line denotes the desired value, unity.

y = 11.571x-1.836 R² = 0.9969

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 5 10 15 20

λs.m

*^0.5

Fr

λs Empirical Determination

0.00 10.00 20.00 30.00 40.00 50.00 60.00

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0 5 10 15 20

m*

∆Pd/∆

Pc

Fr

Accuracy of Method

Fr

m*


Figure 3 shows clearly that the values for pressure forward calculated with λs are about a

maximum of 15% different. This is better than expected (a deviation of 20% is common), and

is acceptable for use in forward calculation (Mills, 2004). Figure 3 also suggests that

increasing mass flow rate and decreasing Fr number seem to adversely affect the accuracy of

the results, but more testing would need be undertaken before any definitive conclusions

about such a relationship are drawn.

Pick up velocity and solids loading ratio.

Figure 3: Solids loading ratio and pick up velocity.

It is worthy of note that an exponential curve fit actually results in a better R2

value (0.901).

However, the results are most likely skewed by the top two outriders in the data, and the

linear curve fit make more intuitive sense in this situation; after all, it would not be expected

that velocity would spike to infinity with decreasing m*, as an infinite velocity is not

physically practical for matter.

From Figure 4, we can see that the approximate pick up velocity required to avoid blockage

is

This relation will return a pick up velocity that is appropriately high enough to prevent

blockage.

y = -0.1687x + 10.99 R² = 0.8623

0

1

2

3

4

5

6

7

8

9

0 10 20 30 40 50 60

Pic

u U

p V

elo

city

(m

/s)

m*

Solids Loading Ratio and Pick up Velocity


Designing Process Having developed a model useful in predicting key system characteristics, it is now possible

to design the proposed system. It should be noted that the processes outlined below are

undertaken in such a way as to minimise operational cost, wear on the system and

maintenance requirements.

System Specifications - Cement meal

- Conveying distance of 750m

- Net required delivery rate of 20 tonne/hr

- 2 bends

- Bend factor 0.5

- 55m of the system is to be vertical.

(see Figure 5 for system schematic)

Figure 4: Schematic of system layout

Assumptions Given that the vertical section of the pipeline accounts for a very small fraction of the

total pipeline, vertical pressure drop can be ignored. This essentially means that the

entire pipe section is considered to be purely horizontal.

The pressure differential across the feeder and pressure drop due to acceleration of the

particles is assumed negligible; both should be minimal for the proper function of a

system (see design of feeder in the report’s recommendations.

It is assumed there is only one pick up destination (the feed bin) and only one delivery

location, at the receiving bin.

Temperature is assumed to be invariant, at 20 degrees Celsius. (isothermal flow)

Incompressible flow is assumed.

It is assumed (with reasonable evidence laid out above) that the forward calculation

model is an adequate indicator of true system parameters.

The pipe need not be finished to a very high tolerance, so a surface roughness of ε

=0.046mm, that of commercial steel is a fair approximation. (Pritchard, 2011)

Design Parameters Ms Ms Ma m* Po Lh D N B R T μ ε

t/hr kg/s kg/s kPa m m J/kg °K °K N.s/m^s m

27 7.778 1.9 2.991 101 750 0.2 2 0.5 286.99 293 1.80E-05 0.000046 Figure 5: Definition of system contents - Red colours are variables


The figure above is easily constructed and forms an effective way of monitoring design

inputs. Red colours were used to designate which constants were to be specified, and so the

values in Figure 6 are not the final chosen. Note that the mass flow rate of solids is variable,

in that a rate differing to the required net transfer may be needed in a batch system.

Method From here, a spread sheet is set up, identical to the one used to test the λs relation derived.

Only the system inputs are different. Again, a guess is needed to start the calculation, so the

same iterative method is used as described in the test results method. Data inputs are this

stage need only be rough estimates in getting the model to work appropriately.

From the use of these spread sheets along with some research into general system

requirements, the particular system specifications are able to be determined, and then the

appropriate system components selected based on these parameters. The process is by no

means linear: given the high level of dependence between variables, the design will require

constant backtracking and cross referencing with previous values to arrive at the optimal

solution.

Stepped Pipelines

Given the likelihood of having to use a stepped pipeline system, similar spread sheets are set

up to calculate a non-uniform bore system. In the case of a single step system, the rough

schematic may be represented as:

Figure 6: Schematic of single step pipeline

Figure 7 is necessary in defining pressures, densities and other variables at different stages

(this diagram is used as the reference system diagram in Appendix F). It is clear, for example,

that section 1-2 of the pipe has its own local Pin and Pout, which in the global nomenclature

become P1 and P2. This means that the pressure, density and velocity at point 2 can be

considered equal (at an instant before and abrupt change). As a result, it is easier to work

backward along the pipe by starting with 2-3; P3 is known as atmospheric, whereas P2cannot

be solved without first defining the section 2-3.From continuity, some values are able to be

set equal to each other, and the total pressure drop is the sum of pressure drops in both pipes.

The aim of stepping the pipeline is to widen pipe area, and subsequently drop velocity:

Increasing area will drop velocity, which can prove beneficial for wear considerations. Other

noteworthy considerations include where the drop occurs along the pipeline, and which side

the bend effects are factored in, both of which have an effect on system performance. (Mills,

2004)


Figure 7: Arbitrary velocity profile for a single step pipeline (only point data evaluated, curves are approximate)

Figure 8 uses arbitrary data to show that in general, if a pipeline is not stepped, the velocity

will continue to increase in the same manner through the pipeline. The figure demonstrates

that stepping reduces the maximum velocity considerably, which will also have positive

effects on the required system pressure drop and on wear in the pipes. (Jones, Mills, &

Agarwal, 2005)

If this velocity profile were still to be unfavourable, the system could be stepped again. This

is achieved by the same process used to introduce the first stepped pipeline. It can be noted

that a generalised velocity profile would shows lower peaks and a more regular velocity

distribution through the pipe. For the velocity profiles of the actual designed system, see

Figures 11 and 12 in the proceeding section.

Running Diagnostics

After the spread sheet has been satisfactorily set up and some data put in, some brief analysis

of results can be undertaken to judge the validity of data. The pick-up velocity used as a

rough lower limit is given by

If the velocity at any point drops below this critical value above, blockage is more likely to

occur. Similarly, care must be taken that the velocity never drops below the inlet velocity of

the first pipe (called the pick-up velocity), as the difference in speeds can cause back pressure

and lead to blockages. As such, a simple ‘IF’ statement in Excel can check if the velocity at

any point in the pipe is more than about 1m/s below the safe velocity, and return an alert for

the designer to warn about potential blockage problems caused by the proposed system.

Ve

loci

ty (

m/s

)

Length through the system (m)

Velocity Profile (Generalised)


Final Design Parameters

Comparison of Stepped and non-stepped systems An important aim of the design process is to minimise wear in the system. This is primarily

achieved by the control of velocities through the pipe: the faster the conveying, the more

abrasive wear occurs (Jones, Mills, & Agarwal, 2005). As such, three different systems were

designed using uniform pipe bore, single stepped pipelines and twice stepped pipelines. The

spread sheets for uniform pipe diameter as well as single and double step systems are

included in the Appendices. Figure 7 above gives the schematic for a single step, while the

Figure below shows the same for a two-step pipeline.

Figure 8: Schematic for the two step, three section pipeline

It is evident that the different arrangements will lead to different performances.

Ma (kg/s)

∆Ptotal (kPa)

Max Velocity (m/s) Power (kW)

Length to 1st Step (m)

Length to 2nd Step (m)

Uniform Bore 1.9 523 57.153 646.155 - -

Single Step 2 386 24.159 592.617 350 -

Double Step 2.2 368 18.989 553.024 250 500 Table 2: Comparison of performance variables of designs using stepped and un-stepped models

It is clear that the system having two steps will provide the optimal specifications (in that it

has lower air requirements, pressure drops, velocities and power usages). The lengths-to-step

were determined again by the guess and check method, with various values being cycled

through to obtain optimal results. From the data in table 2, it is clear a two-step system will

be used in design.

For further graphical comparison, Figure 10 is supplied. The red line represents the predicted

safe velocity Ctest from the results of the test data.


Figure 9: Comparison of Velocity Profiles for Uniform bore, single step and double step systems.

The figure above compares the different profiles in velocity for the designed systems with

uniform bore, then one and two steps. Note that in the two step system, the peaks are closer to

the mean value (i.e. variation is more uniform across the length). This indicates that again,

velocity will be preferable in the two step system.

System Characteristics All subsequent calculations are based on the choice of a 2 step pipeline.

Pipe Bore Given the fact that cement meal is not exceptionally abrasive and will not excessively load

the pipeline, Schedule 40 pipe is a reasonable specification for pipe type (it is a standard

structural steel type). Figure 13 (Saylor.org, 2013) shows standard inner diameters used.

Figure 10: Data sheet used to obtain standard pipe sizes – red boxes denote inner diameters used in spread sheets

With reference to Figure 11, the pipes used in the system will be

Section Nominal Bore Schedule

A 8 inch 40

B 10 inch 40

C 14 inch 40 Table 3: Summary of pipes used in stepped system

0

10

20

30

40

50

60

0 100 200 300 400 500 600 700 800

Ve

loci

ty (

m/s

)

Length along Pipe (m)

Comparison Of Velocity Profiles


These values were arrived at by an iterative process of guessing and checking which standard

pipe sizes yield the most favourable system outputs.

Air Flow rate Using the nominal bores above, values for air flow rate can be put into the spread sheet, and

then checked for validity against Ctest. If the spread sheet returns that a blockage occurs, a

slightly higher value for flow rate is guessed. As a lower mass flow rate of air is optimal in

this system (to reduce power requirements and wear on equipment), guesses were repeated

until the lowest air flow rate that did not cause blockage was found. Ultimately

In volumetric terms

Peak Solids flow rate In such a large system it is highly likely that large pressures will result (given the long

conveying distance). This means that continuous solids conveying of 20 t/hr may be hard to

achieve, and a twin feeder system will most likely need be considered. It can be seen in

Figure 14 that a batch system will lead to a smaller net solid flow rate due to the down time

between batches. As such, twin discharge feeders will need to be used, which will act to

discharge and then refill in offset cycles so a more continuous rate can be achieved.

Figure 12: Variation of solid flow rate in batch conveying

Figure 12 (Mills, 2004) shows that the equivalent continuous flow rate for such a batch

system is about half of the peak (steady state) rate, which is the design parameter. It is

reasonable to assume that a twin discharge system like the one being considered will be

similar, only the rate will not actually drop to zero. As such, the mean delivery rate will shift

up from about 50% of the maximum to 75% (a conservative estimate). Mathematically

Therefore, for a net delivery of 20

, the system should use a steady state delivery of

in order to compensate for the loss of delivery from system downtime.

Solids Loading Ratio Given the specified flow rate of solids and fluid, m* = 3.947.


Pressure Drop If the flow rate of air is set to the designated 1.9kg/s, pressure drop for the twice stepped

system selected is shown by Appendix G to be 368 kPa.

The most favourable pressure drop is provided by the 2 step system (Refer to Table 2). To be

conservative and to account for possible increases in pressure caused by unforseen issues in

running the system, a pressure requirement of 400kPa is a safer stipulation for the system.

This equates to a 4 bar pressure drop.

Power Theoretical power can be calculated from Equation 8 in Appendix A. This calculates power

based on a pressure drop and properties of the air flowing, but not of the solids properties. As

such, a reliable adjustment to get a more conservative estimate of power is to double the

previous calculation (Jones, Mills, & Agarwal, 2005), such that

As such, for the designed system

This is the power required for conveying. It should be noted that Pin needed to be converted to

gauge pressure before calculation, and STP conditions are assumed.


Recommendations

General Given the selected air flow rate and solids flow rate result in an m* value of around 4, the

conveying mode is definitively dilute phase. Indeed, this makes the most sense, given the fact

that over long distances dense phase flow becomes harder to achieve than dilute phase (Jones,

Mills, & Agarwal, 2005). This is further corroborated by the fact that typical dilute phase

velocities range from 10 to 30 m/s, a window in which the designed system operates (Mills,

2004).

A positive pressure design is used, as negative pressure systems are limited to around 1 bar

pressure drop (Mills, 2004), which is well below the demands of the design. It is of note that

cement meal is not typically combustible, so no real design needs to be amended to manage

this problem. Lastly, it is clear that an assembly of twin feeders will be an optimal design

feature to better achieve the specified net delivery rate, because of the ability to continuously

convey.

Air Mover When considering air movers in pneumatic conveying, capacity is often quoted as volumetric

flow rate. As such, a suitable air mover for this system will be capable of supplying the

required 4 Bar of pressure drop and volumetric flow rate of air or 95 m3/min.

The use of a fan is almost immediately ruled out: fans are more effective over smaller

conveying distances, and don’t respond well to higher pressure drops (often material can fall

out of suspension and cause blockage) (Mills, 2004). A number of air movers can be

discounted for their inability to provide the volumetric flow rate required: Liquid ring pumps

provide anywhere between 1 and 70 m3/min, and sliding vane rotary compressors provide a

maximum of 50 m3/min. Both fall short of the 95 m3/min needed for the system (Mills,

2004).

Screw compressors ultimately prove most suitable for the system’s applications, as they are

generally more suited to higher volumetric output: depending on the machine, capacity

ranges from 4 to 700 m3/min. They are also able to handle higher pressures, from about 4 to

9 bar, and are free from pressure pulsations which could cause problems with increased and

unpredictable wear on the pipe (Mills, 2004). As such, a screw compressor is the choice of air

mover for this system.

Possible Component

The HCS Oil free screw compressor produced by Hibon is a possible component for this

systems application (Hibon, 2013). This choice would be dependent of multiple quotes from

different companies.

Figure 113: Suitability of the selected component

Volumetric Flow rate (m^3/min)

Pressure (bar)

System Requirement 95 4

Component Capability 140 3 to 5


Figure 12: Image of a possible screw compressor (Hibon, 2013)

There were a number of options that had max capabilities close to the predicted

characteristics (Atlas Copco, 2013). Choosing such a component could mean that any

subsequent loss of output in the component from age, extended use or wear could drop the

flow rate or pressure below that required. This would likely cause pipe blockage and require

costly hold ups, fault checking and pipe maintenance. Similarly, running a component at its

maximum capacity for extended periods would likely shorten the lifetime and thus increase

the maintenance needs of the component, or even result in the need for premature

replacement. The selected component meets the demand of the designed conveying system

with reasonable overshoot, and so has a reasonable factor of safety in its use. The product

was also chosen as it is oil free; although being more expensive, this type of system provides

better lubrication with less required maintenance and upkeep of the air mover.

Feeder A number of options exist for the type of feeder used in the design. Ultimately it is the

required pressure drop that defines the feeder type selected.

Figure 15: Guide for feeder selection (Mills, 2004)

The only two feeders shown in Figure 15 that are capable of supplying the required 4 bar

pressure drop are Blow tanks and Rotary Valves; the rest can be immediately dismissed as

unable to sustain high enough pressure. Of the two selections, it is worth noting that rotary

valves have significant pressure losses across their openings (which will detract from the

pressure available for conveying), and are especially susceptible to wear (Mills, 2004).

Blow tanks are the best option for feeding this system for a number of reasons

They have no moving parts, meaning lowered need for maintenance and upkeep for

performance

The lack of moving parts also means Blow tanks are resistant to wear, meaning

degradation of the device is minimised, as are maintenance requirements

A single blow tank is able to convey a very wide range of material densities without

modification

Blow tanks are able to hold large batch sizes

Flow rate is easily controlled by simply diverting a proportion of the air flow through

the tank

(Jones, Mills, & Agarwal, 2005)


The only real draw back with blow tanks are their large space occupation, but this is a

reasonable trade off. The best option for a blow tank in this system is to use a top discharge

variety, with a fluidising membrane and valved vent line. There are a number of reasons for

this:

Bottom discharge tanks are not as effective at fluidising the material, meaning

conveying is not as easy (especially for such fine materials as cement meal)

Less room is needed for top discharging blow tanks, meaning a twin blow tank system

is easy to set up

A fluidising membrane is less expensive to fit in the top discharge tank, as they don’t

need to be conical

Added air flows are not needed: a line of conveying air flow is diverted to fluidise the

material, eventually joining the conveying line with the main air flow with no air

losses.

The vent line allows for fast depressurisation of the tank

Some minor draw backs include

The inability to completely discharge material (not of real concern as cement meal is

not really degenerative)

Pressure losses associated with the discharge line (Figure 16) need be mitigated. To

counter this, the shortest possible line with the least amount of vertical section are to

be used to feed the conveying line

The fluidising membrane must be cleaned and maintained to ensure proper

fluidisation occurs

Figure 13: Schematic on desirable Blow tank arrangement for the system (Mills, 2004)

Possible Component

Kockums Bulk Systems Pty Ltd’s standard top discharge Blow Tanks are a possible

component for this system, an example of which is pictured in Figure 17 over-page

(Kockums Pty Ltd, 2012). This choice would be dependent of multiple quotes from different

companies.


Figure 17: A similar Blow tank manufactured by Kockums Bulk Systems

This tank is advertised as being able to deliver a maximum of 200 tonnes of material an hour

(very large safety factor so they machine can be operated at low repetitions, increasing its

lifetime). It is able to be placed in series, and would be available for sizing to order

specification. If the density of loose poured cement is 900-1000 kg/m^3 (Brabender

Technologie, 2012), and two 4 cubic meter tanks are considered:

A capacity of 7.2 tonnes means that in order to attain a solids delivery rate of 20 t/hr, around

three cycles of filling and discharging both tanks are required per hour. This is a reasonably

low number of cycles, and the use of these tanks will therefore minimise the number of

cycles and thus the wear and maintenance requirements on the blow tanks.

Gas-Solid Separation Some means of gas-solid separation is needed to recover the maximum amount of material

possible from the main pipe flow. Also, gas-solid separation is the primary means of avoiding

the polluting of air with particulates in the conveying process. Evidently, the separation will

need to occur at the end of the conveying line, notionally at the end of the receiving hopper.

As a rule of thumb, finer particles lead to a higher cost associated with their separation from

air. Figure 18 shows cement dust is not overly fine, and so will not require exorbitant

amounts of effort or cost to separate.

Figure 14: Comparison of mean particle sizes for some common material (Mills, 2004)

The method of separation is largely dependent on the mean particle size. Cyclones are often

employed to swirl the gas-solids mix around in a vortex, leading to separation. However, this


is effective only for particles >10μm. As Figure 18 indicates, cement meal is therefore too

fine to rely on cyclonic separation. (Mills, 2004)

The best option for gas-solids separation is therefore a filter. A filter acts to trap particles, but

without regular cleaning, will block further airflow and increases the pressure loss. This

means the filter needs to be flushed occasionally to clear it. In a batch system, cleaning can

be achieved by shaking the filter, but given this system uses continuous conveying, time

cannot be taken to shake free particulates (Jones, Mills, & Agarwal, 2005). As such reverse

air pulsing is an inexpensive and easy alternative, and can be carried out while the system is

still running.

The type of material used is for the filter a trade-off between maintenance requirements and

costs. A fabric filters are a very cheap option, but are not especially resilient to chemical

degradation or especially coarse particles, and have a relatively short lifespan. As a more

expensive but longer lasting alternative, glass fibre meshes or other manmade fibre matrixes

can be used (Mills, 2004).

Possible Component

In this system, the solids are not particularly abrasive, and are chemically inert, so the use of

an expensive fibre woven filter is not necessary. It is far more viable to use a cheaper but

equally effective replaceable cloth filter, coupled with a reverse air pulsing system supplied

by an external source of air. The extra maintenance required with occasionally replacing the

cloth filter is negligible for the savings made on equally effective cloth variety.

Maintenance The maintenance requirements of the system has been minimised in the design stage by the

following measures (outlined individually in sections above):

Use of a low-as-possible conveying velocity to reduce wear on pipes

Minimisation of the pressure drop, which again would decrease wear on components,

and thus, maintenance requirements.

Use of a blow tank, which has no moving parts, to eliminate difficult internal

maintenance operations

The introduction of a blow tank valve, fluidising membrane and filter as perishable

parts; these components are easier to replace than others (such as pipe sections) so are

intended to fail before more extensive repairs are needed

Use of a conservative safety factor when selecting components such as the screw

compressor and blow tank. This prevented the possibility of overloading the

component, or having the component operate at full capacity for extended periods,

which would result in greater maintenance costs.

The design of a filter to be automatically pulsated by reversed air eliminated the need

for the filter to be shaken manually.

The proceeding section makes further recommendations for the continual maintenance of the

system.

Potential need for air drying: Cement will tend to react with water in the fluidising

membranes of the blow tanks and on the walls of the pipelines, both of which increasing the

pressure drop. It is therefore important to limit the amount of water entering the system with

the air from the atmosphere. If the system is to be operated in a dry area, this is not of major


concern, but the use of this design in a humid or tropical area may introduce excess moisture.

To avoid the possible blockages and maintenance requirements incurred by this, a desiccant

drying technique should be used. This requires only that the desiccant bed over which air

flows be replaced occasionally. Care should be taken that the desiccant is appropriately

selected to remove sufficient water, and should be able to be recycled for continued use.

Pressure observation points: to aid in the maintenance process, it is suggested that multiple

pressure gauges be placed around the system in key areas. This will allow the operations to

be monitored and any anomalies in performance isolated with relative ease. The location of

these pressure gauges should be:

Before and after the pipe that draws material out of the top discharge blow tank

Before the cloth filter (to ensure excessive clogging is avoided and to time cleaning)

Before and after the fluidising membrane in the blow tank, to check for clogging.

At the point of entry of the solids to the air and at the end of the line (for total loss

measurement)

Monitoring of velocities: as increased wear is associated with higher velocities, the velocity

should be monitored actively at key points (such as the inlet, at the steps and at the exit) to

ensure values close to that of design are being recorded.

Pipe section and bend replacement: the pipes will eventually degrade, with the bends most

likely to suffer the worst wear. Regular checking of pipes should occur so finishing

roughness ε doesn’t change drastically, causing larger pressure drops. Replacement of pipes

is also important in ensuring the system acts as the design intends.


Conclusion Starting with a batch of test data, statistical relations between variables in the system were

able to be derived. After checking the validity of the derived model, a design model for a

system similar to the subject of the test data could be built. The design process in this

instance was not a linear procedure. More often than not it was an iterative approach;

numbers were guessed, iterated, changed in accordance with industry standard then finalised

to define key system components. The choice for each component was justified with a

mixture of literature review and results from the design data.

From this process, the system design is to meet the following specifications:

An overall pressure drop of 368kPa (3.68 bar)

A power consumption of 553kW

A steady state flow rate of cement meal of 27 tonnes/hr

A flow rate of air or 1.9kg/s

m* = 3.947

Positive displacement system for conveying

Dilute phase conveying

Schedule 40 pipes used

Pipeline stepped twice. First at 250m along the horizontal section from an 8 inch

nominal bore to 10 inch, then once more another 250m along the system (at the base

of the vertical rise) from 10 to 14 inch nominal bore.

Twin 4 m3 top discharge blow tanks in series (brand suggestion made)

Continuous conveying from the twin blow tanks.

Fluidising membranes fitted to the blow tanks

Use of a separate line and valve in depressurising the blow tanks

Use of supplementary air from air mover to fluidise blow tank material.

Oil free screw compressor (brand suggestion made made) as an air mover

A removable cloth filter fitted on the receiving bin for gas-solid separation

An air pulse jet set up to reverse cleanse the filter

Potential addition of desiccant bed at air inlet for water vapour removal

Pressure gauge and velocity gauge at key parts of the system

This design will be an effective means of conveying cement meal as required, and aims to

minimise wear and maintenance upkeep.


Bibliography Atlas Copco. (2013). GA 315-500: Oil-injected rotary screw compressors. Retrieved August

25, 2013, from Atlas Copco Site:

http://www.atlascopco.com.au/auus/products/Product.aspx?id=1524396&productgrou

pid=1473343

Brabender Technologie. (2012). Standard Material Densities. Retrieved August 25, 2013,

from Conveying Feeders:

http://www.lossinweightfeeder.com/standards/documents/1500-C01-2.pdf

Hibon. (2013). Hibon Product Site. Retrieved August 25, 2013, from Blowers, Compressors

and high Vaccum Packages: http://www.hibon.com/IS/Product.aspx-eu_en-22772

Jones, M., Mills, D., & Agarwal, V. K. (2005). Handbook of Pneumatic Conveying

Engineering. New York: Marcel Dekker.

Kockums Pty Ltd. (2012). Pneumatic Conveying Systems. Retrieved August 20, 2013, from

Kockums Bulk Systems:

http://www.kockumsbulk.com.au/system/powder_handling_product/pdf/0000/0020/D

ense-Phase-Conveyor.pdf

Kockums Pty Ltd. (2013). Power Handling Products. Retrieved August 20, 2013, from

Kockums Bulk Systems:

http://www.kockumsbulk.com.au/powder_handling_products/20-

dense_phase_pneumatic_conveyor

Mills, D. (2004). Pnuematic Conveying Design Guide. Oxford: Elsevier.

Pritchard, P. J. (2011). Fox and McDonald's Introduction to Fluid Mechanics. New York:

Wiley.

Saylor.org. (2013). Schedule Pipe Sizes. Retrieved August 20, 2013, from Saylor: Free

Educator: http://www.saylor.org/site/wp-content/uploads/2011/07/ME303-4.1.1.pdf

Streeter, V. L., & Wylie, B. E. (1983). Fluid Mechanics. New York: McGraw-Hill Co.


Appendix A – Equations 1.

2.

3.

4.

5.

6.

7. =

8.

9.

10.

11. =

12.

13.

√

14. S*=

15.

16.

- is solids loading ratio, is the mass

flow rate of fluid (s for solid)

- P is pressure, is pressure drop

- is density, x denotes which point

density is taken at, R is ideal gas constant

and T is absolute temperature (°K)

- C is the velocity at point x, d is diameter

- Re is Reynold’s number, is dynamic

viscosity

- is the fluid friction factor, is surface

roughness

-L is the length of the system

-N is the number of bends, B is the bend

factor

- is the solids friction factor

- Fr is the Froude’s number, g is

gravitational constant (9.81 m/ )

- S* has no real interpretation, but was

useful to define for plotting purposes

- B and c are constants


Appendix B – Test Data

Test Ma Ms ∆P m* Po Pavg ρi ρo ρavg Ci Co Cavg Re λf ∆Pair ∆Pbend λs Fr b*

kg/s kg/s kPa kPa kPa kg/m^3 kg/m^3 kg/m^3 m/s m/s m/s kPa kPa

1 0.081 2.120 315 26.173 101 258.5 4.947 1.201 3.074 7.421 30.567 11.943 108105 0.02170 15.797 41.702 0.01351 16.563 0.069

2 0.081 2.380 348 29.383 101 275.0 5.340 1.201 3.270 6.876 30.567 11.226 108105 0.02170 14.849 43.831 0.01439 15.569 0.078 3 0.050 1.480 249 29.600 101 225.5 4.162 1.201 2.682 5.445 18.868 8.451 66732 0.02288 7.277 20.513 0.02350 11.720 0.128 4 0.047 1.560 257 33.191 101 229.5 4.257 1.201 2.729 5.004 17.736 7.806 62728 0.02306 6.368 19.900 0.02518 10.825 0.145 5 0.042 1.490 247 35.476 101 224.5 4.139 1.201 2.670 4.600 15.849 7.131 56055 0.02341 5.276 17.330 0.02806 9.889 0.167 6 0.038 1.530 259 40.263 101 230.5 4.281 1.201 2.741 4.023 14.340 6.283 50716 0.02373 4.265 15.630 0.03305 8.714 0.210 7 0.050 1.550 260 31.000 101 231.0 4.293 1.201 2.747 5.279 18.868 8.250 66732 0.02288 7.104 20.941 0.02410 11.441 0.134 8 0.034 1.550 273 45.588 101 237.5 4.448 1.201 2.824 3.465 12.831 5.456 45377 0.02412 3.368 13.711 0.04021 7.567 0.271

9 0.030 1.500 275 50.000 101 238.5 4.472 1.201 2.836 3.041 11.321 4.794 40039 0.02459 2.661 11.637 0.04817 6.649 0.341 10 0.026 1.300 277 50.000 101 239.5 4.495 1.201 2.848 2.622 9.812 4.138 34700 0.02516 2.037 8.704 0.06577 5.738 0.465 11 0.050 1.580 267 31.600 101 234.5 4.376 1.201 2.789 5.179 18.868 8.127 66732 0.02288 6.998 21.015 0.02473 11.270 0.139 12 0.022 1.100 298 50.000 101 250.0 4.745 1.201 2.973 2.102 8.302 3.354 29362 0.02589 1.438 5.970 0.10465 4.652 0.740 13 0.049 1.860 285 37.959 101 243.5 4.590 1.201 2.896 4.838 18.491 7.670 65397 0.02294 6.489 23.228 0.02378 10.637 0.146 14 0.032 1.750 297 54.688 101 249.5 4.733 1.201 2.967 3.064 12.076 4.888 42708 0.02434 2.866 13.820 0.04354 6.779 0.322 15 0.019 0.890 259 46.842 101 230.5 4.281 1.201 2.741 2.012 7.170 3.142 25358 0.02658 1.194 4.531 0.12036 4.357 0.824

Table 4: Test Data spreadsheet. All calculations are referenced above.


Appendix C – Forward Calculation

Table 5: Spreadsheet for testing of forward calculating solids friction factor

Ma Ms ∆Pguess m* Pavg ρi ρavg Ci Co Cavg Re λf Fr λs ∆Pair ∆Pfrict ∆Pbend ∆Pcalc ∆Pd/∆Pc

kg/s kg/s kPa kPa kg/m^3 kg/m^3 m/s m/s m/s kPa Pa kPa kPa data/calc

0.081 2.120 303 26.17 252.49 4.80 3.00 7.64 30.57 12.23 108105 0.02170 16.957 0.01251 16.17 244.10 42.69 303 1.0397 0.081 2.380 331 29.38 266.50 5.14 3.17 7.15 30.57 11.58 108105 0.02170 16.066 0.01304 15.32 270.59 45.23 331 1.0509 0.050 1.480 245 29.60 223.50 4.11 2.66 5.51 18.87 8.53 66732 0.02288 11.825 0.02281 7.34 216.60 20.70 245 1.0178 0.047 1.560 259 33.19 230.50 4.28 2.74 4.98 17.74 7.77 62728 0.02306 10.778 0.02553 6.34 232.98 19.81 259 0.9918 0.042 1.490 258 35.48 230.00 4.27 2.74 4.46 15.85 6.96 56055 0.02341 9.652 0.03024 5.15 236.03 16.92 258 0.9570 0.038 1.530 273 40.26 237.50 4.45 2.82 3.87 14.34 6.10 50716 0.02373 8.457 0.03618 4.14 254.08 15.17 273 0.9474 0.050 1.550 254 31.00 228.00 4.22 2.71 5.37 18.87 8.36 66732 0.02288 11.592 0.02312 7.20 225.39 21.22 254 1.0244 0.034 1.550 289 45.59 245.50 4.64 2.92 3.32 12.83 5.28 45377 0.02412 7.321 0.04432 3.26 272.93 13.26 289 0.9431

0.030 1.500 298 50.00 250.00 4.75 2.97 2.87 11.32 4.57 40039 0.02459 6.343 0.05506 2.54 284.31 11.10 298 0.9230 0.026 1.300 280 50.00 241.00 4.53 2.87 2.60 9.81 4.11 34700 0.02516 5.703 0.06695 2.02 269.33 8.65 280 0.9893 0.050 1.580 258 31.60 230.00 4.27 2.74 5.31 18.87 8.29 66732 0.02288 11.491 0.02327 7.13 229.23 21.43 258 1.0357 0.022 1.100 261 50.00 231.50 4.31 2.75 2.32 8.30 3.62 29362 0.02589 5.023 0.08450 1.55 253.39 6.45 261 1.1401 0.049 1.860 297 37.96 249.50 4.73 2.97 4.69 18.49 7.49 65397 0.02294 10.381 0.02558 6.33 268.03 22.67 297 0.9595 0.032 1.750 334 54.69 268.00 5.17 3.19 2.80 12.08 4.55 42708 0.02434 6.311 0.05314 2.67 318.49 12.87 334 0.8892 0.019 0.890 234 46.84 218.00 3.98 2.59 2.16 7.17 3.32 25358 0.02658 4.607 0.10234 1.26 227.70 4.79 234 1.1080


Appendix D – System Parameters Table

This table is replicated above every design spreadsheet, containing system parameters. Red font indicates a design variable.

Appendix E – Uniform Pipe Bore Design Spread sheet

Design Parameters Ms Ms Ma m* Po Lh D N B R T μ ε

t/hr kg/s kg/s kPa m m J/kg °K °K N.s/m^s m

27 7.7778 1.900 4.0936 101 750 0.2027 2 0.5 286.987 293 1.80E-05 0.000046

Design characteristics ∆Pguess Pavg ρi ρo ρavg Ci Ci Test Co Cavg Re λf Fr λs ∆Pair ∆Pbends ∆Pfrict

kPa kPa kg/m3 kg/m3 kg/m3 m/s m/s m/s m/s kPa kPa kPa

478.15 340.0

76 6.887531

752 1.2011

36 4.044

33 9.96

7 10.4096 57.152842 16.97

4 7703

87 0.0195 12.05

79 0.0648

39 42.20 2.57 478.15

RESULTS

Max C

Power (Rough)

Practical Power ∆Ptot Ctest Check

57.1 323.078 646.155 523

10.613

No Blockage

m/s kW kW kPa m/s


Appendix F – Single Step Design Spread sheet

Design characteristics (part 2)

Lh D ∆P Pin=P2 Pout =

P3 Pavg ρi=ρ2 ρo=ρ3 ρavg Ci=C2 Co=C3 Cavg Re λf Fr λs ∆Pair ∆Pbend ∆Pfrict ∆P

m m kPa kpa kpa kPa kg/m^3 kg/m^3 kg/m^3 m/s m/s m/s kPa kPa kPa kPa

350 0.307 107 208 101 154.507 2.474 1.201 1.837 10.369 21.356 13.96 437634 0.0198 8.043 0.1267 4.05 0.88579 102.08 107

Design characteristics (part 1)

Lh D ∆P Pin=P1 Pout =

P2 Pavg ρi=ρ1 ρo=ρ2 ρavg Ci=C1 Co=C2 Cavg Re λf Fr λs ∆Pair ∆Pbend ∆Pfrict ∆P

m m kPa kpa kpa kPa kg/m^3 kg/m^3 kg/m^3 m/s m/s m/s kPa kPa kPa kPa

400 0.202 293 501 208 354.688 5.962 2.474 4.218 9.914 24.159 14.01 664348 0.0196 9.948 0.0858 16 0 277.31 293

RESULTS

Max C Power

(Rough) Practical Power ∆Ptotal Ctest A B

24.159 296.31 592.617 386 10.647 No

Blockage No

Blockage

m/s kW kW kPa m/s


Appendix G – Two Step Design spread sheet

RESULTS Max C Power (Rough) Practical Power ∆Ptotal Ctest A B c

18.989 276.512 553.024 368 10.31887 No Blockage No Blockage No Blockage

m/s kW kW kPa m/s

Design characteristics (part C) Lh D ∆P Pin=P3 Pout = P4 Pavg ρi=ρ3 ρo=ρ4 ρavg Ci=C3 Co=C4 Cavg Re λf Fr λs ∆Pair ∆Pbends ∆Pfrict ∆P

m m kPa kpa kpa kPa kg/m3 kg/m

3 kg/m

3 m/s m/s m/s kPa kPa kPa kPa

250 0.333 65 166 101 133.5 1.975 1.201 1.588 11.04 18.16 13.73 40359 0.01990 7.601 0.1405 2.24 0.371 62.429 65

Design characteristics (part B) Lh D ∆P Pin=P2 Pout = P3 Pavg ρi=ρ1 ρo=ρ2 ρavg Ci=C1 Co=C2 Cavg Re λf Fr λs ∆Pair ∆Pbends ∆Pfrict ∆P


3 kg/m


250 0.254 113 279 166 222.3 3.315 1.975 2.645 11.31 18.98 14.17 52912 0.01971 8.982 0.1034 5.15 0.658 106.86 113

Design characteristics (part A)

Lh D ∆P Pin=P1 Pout = P2 Pavg ρi=ρ1 ρo=ρ2 ρavg Ci=C1 Co=C2 Cavg Re λf Fr λs ∆Pair ∆Pbends ∆Pfrict ∆P


3 kg/m


250 0.203 190 469 279 373.9 5.579 3.315 4.447 10.55 17.76 13.24 66303 0.01958 9.389 0.0953 9.41 0 180.99 190

Pneumatic Conveying Assignment (Final)

Documents

Transcript of Pneumatic Conveying Assignment (Final)