University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Chalmers University of Technology | results regarding higher reduction ratio at lower CSS as the electrical energy in the motor, via
kinetic energy, is transformed to surface energy when the rocks break.
Average Power draw vs. CSS
140
]W
k120
[
w a rD100 y = -1,5462x + 181,82
re
80 R² = 0,9837
w
o 60
P
e 40
g
a
re 20
v
A
0
32 34 36 38 40 42 44 46 48 50 52
CSS [mm]
Figure 34 - The average power draw measured for the five CSS settings displaying a negative trend when increasing
the close side setting
4.3 Hydrostatic pressure
The pressure signals sampled at 500 Hz for 120 seconds are plotted in Figure 36. The same
argument as in the case with the power draw signals can be made for the pressure signal. The
signal has very high amplitude and appears to be noisy. However as previously described the
amplitude is not a result of noise but due to cyclic behaviour in the crusher and super positioned
discrete crushing events. A detailed view of two of the signals for one second of operation can
be seen in Figure 35. The data reveals that the frequency of the variation matches the eccentric
speed of the crusher. Hence this provides yet further evidence that the crusher is not fed in a
satisfactory manner or there may be uneven wear on the concave.
Pressure vs. time for two CSS
3,5
3
]a
2,5
P
M 2
[
e
ru
s1,5
s
e
rP
1
0,5
0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
Time [s]
Pressure_34 Pressure_50
Figure 35 - In order to get a closer look at the measured pressure signal one second of operation is displayed in the
plot for two of the CSS settings. The time span between each peak corresponds to the eccentric speed of the mantle.
35 |
Chalmers University of Technology | 5. ROCK MATERIAL MODEL DEVELOPMENT
The aim of this section is to:
- Describe the process of building a rock material model in DEM
- Present the experiments and simulations performed in order to calibrate the rock model
Five different 3D scanned rock geometries with random shapes have been chosen as models for
the meta-particles. When building a rock population in the DEM environment it is necessary to
cut the particle size distribution at some level. Here the smallest size class was chosen to be 31.5
mm. The total population consists of six different types of particles all with different size and
shape. During the development of the feed population packing clusters were made for the six
different sizes. Unfortunately only the two largest cluster sizes were later used in the DEM
simulations due to practical problems with the particle replacement procedure.
A substantial amount of time was spent finding the optimum bi-modal size distribution for each
size class. When designing the meta-particles a few aspects need to be balanced. Generally the
quality of the cluster stands against the computational load a meta-particle will place on the
overall system. High quality is generally achieved by having a high number of small fraction
particles. This leads to an optimization problem as a high number of particles in each meta-
particle reduce the total number of meta-particles possible to use in the total feed population.
A number of different particle beds have been tested iteratively in order to find a suitable bi-
modal size distribution that complies both with breakage quality whilst minimizing the number
of particles. The key finding/idea that led to a breakthrough during these development efforts
was to use relatively large top size in the bi-modal distribution. By having a large gap between
the large and small particles a high packing density can be obtained while keeping the number
of particles down. Also it results in a bonding structure that is complex and gives heterogenic
behaviour. It is important to emphasize that the ambition with the BPM model is not to capture
the intricate complexity of rock mechanics but more to build a rock particle that responds with
similar behaviour when interacting with the crusher. As long as the particle fractures at similar
force levels and breaks apart in progenies it is expected that it can be used for the stated
purposes in the thesis.
5.1 Single particle breakage experiment
In order to calibrate the BPM model a large number of Single Particle Breakage (SPB) tests
have been conducted on the sample material. In each test the particle is weighed and placed in
the hydraulic compression test rig. The particle is then compressed 10% which normally is
enough to break the rock. Particles commonly break in two different ways. When loaded a
stress is built up in the particle body until reaching a high enough level to cause crack
propagation through the particle in a single main fracture. The other case can be characterized
by sequential fracture where e.g. a sharp corner is chipped off and the particle is reloaded with
subsequent local damage as a result. The breakage mode is mostly a matter of loading condition
as previously described in Table 1. Prior to breakage tests all particles have been weighed in
order to test the hypothesis of a correlation between particle strength and particle mass. The
LabVIEW software used for logging the data from the compression rig automatically registers
the distance between the compacting plates when experiencing 50 N of load. This gives an
estimate of the particle size. The particles are then semi-continuously, as a manual hydraulic
pump is used, loaded until broken or reaching 10% compression. From the force-displacement
curves the critical force for the main breakage can be logged. The critical force values have been
plotted against rock height in Figure 40 and against rock mass in Figure 41. The graphs show
that the strength is highly scattered in both plots. Some potential reasons for this are listed
below:
38 |
Chalmers University of Technology | Stochastic orientation of particles
Anisotropic structure and strength
Particle shape difference and hence variation in loading condition which means totally
different stress fields within the rock body.
Stochastic variation in the rock strength
Variation in the compression movement due to manual hydraulic pump
Variation due to robustness problems with the compression rig frame
Figure 39 shows the force displacement curves for 7 of the 116 performed SPB tests. As can be
seen there is a staircase pattern due to the use of a manual hydraulic pump. Ideally these tests
should be performed using a compressive rig with a motorized hydraulic pump where the rate
can be controlled. As can be seen in the plot the force is built up until the particle breaks. The
peak force at which breakage occurs is defined as the critical force.
SPB experiments -Force vs. Displacement
12000
10000
8000
S1
]N S2
[ ecroF 6000 S S3
4
S5
S6
4000 S7
2000
0
0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%
Compression ratio [%]
Figure 39 - The figure shows the force-displacement plot for 7 of the 116 performed SPB tests. The staircase pattern
seen is related to the manual hydraulic pumping motion.
Peak Force -Rock Height
70000
60000
50000 y = 18,817x2-339,69x + 6793,1
R² = 0,5609
Granite
40000
]N
[
cF30000
R² =
R 0²
,3
=
1
50 4,5278
A
L
Pim
on le
yp
a
.h
r
(i Gb (Go rali rt nae
in teit )e)
Linear (Amphibolite)
20000 y = -7,6805x2+ 1119,8x -13076 Poly. (Amphibolite)
10000
0
10 15 20 25 30 35 40 45 50 55 60
Rock height [mm]
Figure 40 - Scatter plot of critical breakage force vs. rock height. The height is defined as the distance between the
plates when the rock is loaded by 50 N.
39 |
Chalmers University of Technology | Peak Force -Rock mass
70000
60000
50000
R² = 0,5039 Granite
40000
]N R² = 0,2546 Amphibolite
[
cF30000 y =
R
6
²
5 =1 0,8 ,56 5x 80, 76327 L Pi on wea er
r
( (G Gr ra an ni it te e)
)
Linear (Amphibolite)
20000 y = 1612x0,4437 Power (Amphibolite)
R² = 0,328
10000
0
0 100 200 300 400 500 600
Rock mass [g]
Figure 41 - Scatter plot of critical breakage force vs. rock mass. Each rock specimen was weighed before being
crushed.
Eq. 19 has been used to calculate the approximate rock strength of all rock samples. The
average strength was calculated to 13.29 MPa.
5.2 BPM calibration
The general objective with the calibration is to find the bond parameters that give relevant
breakage behaviour similar to the experimental results. The general process for creating meta-
particle clusters has been previously described in the Methods chapter. When beginning to put
together different cluster particle structures it was found that it is not feasible to use the same
fraction size distribution for all the size classes. The biggest size class particle can be fitted with
relatively large fraction particles, in fact bigger than the smallest size class. Hence two different
bi-modal fraction particle beds were created, one fine and one coarse. This was done in order to
minimize the number of particles used for each meta-particle. Onwards the two bed
configurations are denoted as BPM and BPM .
fine coarse
5.2.1 Calibration simulations of BPM
fine
In the case of the BPM the meta-particle representing the smallest size class was chosen due
fine
to its close resemblance in shape and size to the particles tested in the single particle breakage
experiments. A 3D rock shape was randomly chosen as subject of breakage, see Figure 42. The
strength parameters simulated for the BPM calibration can be seen in Table 4. The start
fine
values have been calculated based on previous work done by Potyondy and Cundall [17].
Table 4 - Strength parameters simulated for the calibration of BPM
fine
Normal stiffness Shear Stiffness Normal Critical Stress Shear Critical stress
Run [GN/m3] [GN/m3] [MPa] [MPa]
1 2500 1000 10 5
2 2500 1000 10 12,5
3 2500 1000 10 20
4 2500 1000 45 5
5 2500 1000 45 12,5
6 2500 1000 45 20
7 2500 1000 80 5
8 2500 1000 80 12,5
9 2500 1000 80 20
40 |
Chalmers University of Technology | Figure 42 - Pictures of the 3D rock model used in the breakage calibration of BPM
fine
In Figure 43 the SPB breakage process can be seen in the DEM environment. The bonds are
displayed instead of the fraction particles as it is then easier to observe the crack propagation.
The simulation shown is no. 6 and, as can be seen in Figure 44, the particle breaks in two
distinct breakage events. The result is a set of progeny particles of various sizes as well as the
free fraction particles broken from the main body.
Figure 43 - Snapshots of the calibration breakage simulation of the BPM . The pictures shown are a time-sequence
fine
from simulation no. 6 in Figure 44.
For the particle shape chosen the critical force that corresponds to 13.29 MPa is approximately
8000 N. Hence simulation no.5 and no.6 would be possible candidates for a relatively good
match.
A lot of tests have been performed in order to reduce the compression ratio at which the
breakage occurs. When comparing to the SPB experiments the compression ratios are around
one magnitude too high. In order to achieve correct compression ratio the particle and bond
stiffness needs to be extremely high. That leads to very small time-steps that are not feasible.
41 |
Chalmers University of Technology | SPB DEM -Force vs. Displacement
14000
12000
]N10000
[
e
c
ro
F 8000
e
v
is
s 6000
e
rp
m
o 4000
C
2000
0
0,0% 5,0% 10,0% 15,0% 20,0%
Compression ratio [%]
1_10x5 [Mpa] 2_10x12,5 [Mpa] 3_10x20 [Mpa]
4_45x5 [Mpa] 5_45x12,5 [Mpa] 6_45x20 [Mpa]
7_80x5 [Mpa] 8_80x12,5 [Mpa] 9_80x20 [Mpa]
Figure 44 - The force-displacement plot for the nine different tested strength combinations simulated for the fine bed
configuration.
5.2.2 Calibration simulations of BPM
coarse
For the BPM a cylinder has been used as geometry model instead of a 3D scanned rock
coarse
particle. This shortens the calibration procedure as it is not needed to use the particle
replacement functionality. Instead a particle cluster structure is created and bonded together
inside a cylinder which is a very simple task to do in EDEM. The same base-file can be used for
all the simulations just by changing the bonding strength parameters. Twenty simulations have
been performed using batch simulation mode and the settings tested can be seen in Table 5.
Table 5 - Strength parameters simulated for the calibration of BPM
coarse
Normal Stiffness Shear Stiffness Normal Critical Shear Critical
Run [GN/m3] [GN/m3] Stress [MPa] stress [MPa]
C1 1670 667 45 16
C2 1670 667 45 12
C3 1670 667 45 8
C4 1670 667 45 4
C5 1670 667 45 2
C6 1670 667 45 1
C7 2004 800.4 45 16
C8 2004 800.4 45 12
C9 2004 800.4 45 8
C10 2004 800.4 45 4
C11 2004 800.4 45 2
C12 2004 800.4 45 1
C13 1670 667 30 30
C14 1670 667 33 27
C15 1670 667 36 24
C16 1670 667 39 21
C17 1670 667 42 18
C18 1670 667 45 15
C19 1670 667 48 12
C20 1670 667 51 9
In Figure 45 the cylindrical particle bed is shown as well as the bonds between fraction particles.
The picture clearly shows the bi-modal size distribution. If examining the bonding network it
42 |
Chalmers University of Technology | can be seen that the large fraction particles have a lot of contacts due to the small particles in
between. Compared to a mono-size distribution this creates a high level of irregularity and
interlocking within the body.
,
Figure 45 - Snapshot from EDEM showing the cylindrical particle bed to the left and the bonds between fraction
particles to the right.
The breakage sequence from one of the calibration simulations can be seen in Figure 46. The
force network through the cylindrical bed can clearly be seen. The forces build up until a high
enough stress results in crack propagation. One of the problems during the calibration was that,
for some of the strength settings, the rock body broke due to local damage only. It was not
possible to build up a stress field in the bond network. The bonds in contact with the
compression plates could not withhold the local stress. These local damage phenomena can be
seen in the fourth picture from the right in Figure 46.
Figure 46 - Snapshots of the calibration breakage simulation of the BPM . The pictures shown are a time-
coarse
sequence from simulation no.C9 in Figure 47.
Figure 47 shows the force-displacement curves for the 20 simulations. It is evident that there is a
large output variation. The settings were chosen iteratively by first testing the calculated start
values according to the Potyondy and Cundall [17] calculations. Then five new settings were
chosen based on the previous results. This iterative process continued until a satisfactory result
was obtained. For the cylindrical body the critical force corresponding to ~13.3 MPa is around
95’000 N. Hence the settings chosen for the crusher simulations correspond to simulation C15.
43 |
Chalmers University of Technology | 6. CRUSHER SIMULATION
The aim of this section is to:
- Present results from the performed DEM simulations
- Display the capability of the DEM environment in terms of visualization and extraction
of data not possible to measure in real operations
Even though DEM simulations incorporating breakage and BPM models have been performed
in previous projects at Chalmers Rock Processing Systems (CRPS) the complexity and ambition
in this work have led to difficulties. The process of building BPM models are explained in
previous chapters and should be relatively straight forward however there are a lot of things
that can go wrong. These problems have occupied a lot of time hence only two out of five
simulations have been performed. When the first simulation (CSS34) was started it was found
that the computation was too slow. This led to scalping of a large section of the feed material in
order to reduce the number of particles.
Table 6 - Simulation parameters for the DEM simulations
Parameter Value Unit
DEM Material properties
Rock Steel
Solid density 2630 7800 [Kg/m3]
Shear stiffness 5.571e8 7e10 [Pa]
Poisson’s ratio 0.35 0.3 [-]
Rock-Rock Rock-Steel
Coefficient of static friction 0.5 0.7 [-]
Coefficient of restitution 0.2 0.25 [-]
Coefficient of rolling friction 0.001 0.001 [-]
BPM parameters
Normal Stiffness 1670 [GN/m3]
Shear Stiffness 667 [GN/m3]
Normal Critical Stress 36 [MPa]
Shear Critical stress 24 [MPa]
Bond disc radius 6.4 [mm]
Machine
Crusher model Svedala H6000 [-]
Chamber type CHX [-]
Eccentric throw 48 [mm]
Eccentric speed 5 [Hz]
Close side setting [34,50] [mm]
Section simulated 1/4 [-]
Liner geometry 3D scanned and modelled [-]
Simulation
Time step 3e-7 [s]
Write out frequency 500 [Hz]
No. of particles (CSS34) 50364 [-]
N0. of particles (CSS50) 92508 [-]
Simulation time (CSS34) 172 [CPUH]
Simulation time (CSS50) 404 [CPUH]
CPU clock freq. 3.33 GHz
CPU cores 8 [-]
Result from the CSS34 simulation was not promising. The particle flow behaviour showed an
unrealistic splattering effect with particles exploding during compression. Also, due to the
reduced feed batch the chamber never reached choke fed condition.
For the CSS50 simulation two things were changed. In order to get rid of the splattering
problem the particle velocity in the simulation was limited to 20m/s and the angular velocity to
45 |
Chalmers University of Technology | 1200 degrees/s. Sometimes particles overlap too much resulting in very high forces hence high
velocities consequently giving a chain reaction. By setting a max velocity in the simulation these
unwanted peaks are removed giving a more stable particle flow. The second measure to
improve the simulation was to use the whole feed batch originally calculated to fill the chamber.
Settings for both simulations can be seen in Table 6. Comments below generally refer to the
CSS50 simulation.
6.1 Particle Flow Behaviour
The flow of particles through the crushing chamber looks realistic compared to observations
from experiments. However it is not possible to make live observations of the particle flow
during crushing in a choke fed condition. To some degree it is possible to observe rocks that
flow and break when the crusher is run empty.
Figure 48 shows meta-particles in the chamber, with each fraction particle coloured by velocity.
From the simulations it can be seen that each time a rock particle is captured between the
mantle and concave the vertical velocity component either reaches zero or sometimes positive
values. The positive velocity component is an effect of the squeezing effect between two planes
at an angle. This suggests that the friction between the rock particle and the liner surface is an
influential factor for the mass flow through the crusher. This also means that the bulk flow
characteristics are important to calibrate in addition to the breakage behaviour. An interesting
phenomenon in the discharge region is that particles just leaving the chamber are pushed or
thrown away radially by the mantle each time it passes. The rock cascade this results in may
explain wear patterns experienced in industrial operations on the walls and supporting beams
below the liners. The observed flow behaviour is also supported by the analytical findings by
Evertsson [31].
Figure 48 - Snapshot from DEM simulation (CSS50). Particles are coloured according to velocity.
46 |
Chalmers University of Technology | 6.2 Breakage behaviour
When the mantle rolls over the particle bed a force pulse can be observed with maxima located
at the current angular position i.e. the closest gap position. As only a section of the crusher is
simulated merely one whole pressure pulse can be observed simultaneously in the particle bed.
If simulating the whole crusher it would be possible to see how the pressure pulse travels
around the crusher following the mantle motion. From the simulations it can be observed that
particles break incrementally due to both interactions between geometries as well as other
particles. The population of meta-particles is not entirely broken hence a product size
distribution is obtained. A positive result is that the meta-particles are able to break up in
progeny particles which are in turn subjected to new breakage events. This shows that the BPM
approach not only captures the breakage but is also able to react to different compressive
breakage modes. This feature is not yet possible to achieve in PBM breakage models. The PBM
approach is based on a threshold force or energy level at which the particle will be replaced with
a set of pre-determined progeny particles. This leads to a number of problematic issues and
scenarios when using the PBM approach;
1. Particles are not able to respond to different compressive modes.
2. Consider a rock particle compressed at ratio X that leads to an initial crack formation
leading to two main progeny particles. The particle may be further compressed in the
same compression event breaking the two progeny particles into a new set of
progenies. This type of sequential breakage would be difficult to handle using the
PBM approach as the breakage is solely governed by a statistical probability function.
3. For the same compression ratio two particles may break up into a distribution of
progeny particles in two distinctly different ways.
Using an impact breakage test as the basis for calculation of threshold values and progeny
distributions is wrong since the breakage modes are completely different.
In Figure 49 the numbers of intact bonds in the two simulations are shown. The staircase
pattern is of course a result of simulating only a 90 degree section. As can be seen bonds breaks
at compression events five times per second. Linear regression gives the bond breakage rate. As
previously mentioned the feed batch to the CSS34 simulation was scalped which can be seen
from the number of bonds in the beginning of the simulation. These results show a higher
breakage rate for CSS50 than CSS34 however this is most likely due to more material in the
chamber. Hence no conclusions should be drawn based on these results.
No. of intact bonds vs. time
300000
250000
[
sd-]
200000
n
o
b
tca150000
tn
i fo y = -80083x + 314861
.o100000 R² = 0,9968
N
y = -52000x + 160190
50000 R² = 0,9823
0
0 0,5 1 1,5 2 2,5 3 3,5 4
Time [s]
CSS 34 - Number of Intact Bonds CSS 50 - Number of Intact Bonds
Linear ( CSS 34 - Number of Intact Bonds) Linear ( CSS 50 - Number of Intact Bonds)
Figure 49 - The plot shows the number of intact particles still in the simulation. The reason for the difference in start
values is due to the different number of meta-particles in the two cases.
47 |
Chalmers University of Technology | In Figure 50 one compression event is shown through a series of pictures. In the first picture
particles are beginning to break. In the third picture the gap is a CSS and a lot of broken bonds
are displayed as black vectors. In the subsequent pictures the mantle rolls over and in the last
pictures the particles are free to fall due to gravity.
Figure 50 - The series of pictures displays one compression event. The meta-particles are shown as bonds colourer
according to perceived normal force.
In order to calculate the pressure from the DEM simulation the force response on the mantle
has been extracted. The total force magnitudes as well as the z-components over time are shown
in Figure 51. A large difference can be observed between the two simulations. The reason is the
lack of feed material in the CSS34 case. Focusing on the CSS50 simulation it shows that a
substantial part of the force between particles and the mantle is accounted for by the z-
component. It would be very interesting to see what the shape of the force response is when
48 |
Chalmers University of Technology | simulating a full crusher. This would give an indication of the natural stochastic behaviour and
variation due to the crushing process. By altering the feed properties and conditions it would
then be possible to investigate the causes of variation in problematic pressure signals.
Force response on mantle vs. time
600000
500000
400000
]N
[
e300000
c
ro
F
200000
100000
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
Time [s]
CSS50 - Total Force (magnitude) CSS50 - Total Force (Z-comp)
CSS34 - Total Force (Z-comp) CSS34 - Total Force (magnitude)
Figure 51 - Force response on the mantle geometry from the particles acting upon it. Both the total force magnitude
as well as the vertical Z-component is shown for both simulated settings.
6.1 Size reduction
Currently it is difficult to extract the product size distribution from the DEM simulations. The
size distribution has to be calculated by identifying remaining clusters of fraction particles, still
bonded together, and estimating their size. There is currently no method or functionality within
EDEM that can collect this information. The following procedure is suggested as an approach
to calculating the particle size distribution from a crushed BPM rock population;
i. Export data
ii. Import dataset to MATLAB
iii. Identify clusters
iv. Remove duplicate clusters
v. Estimate cluster size
vi. Summarize all particles not in clusters
vii. Estimate the size distribution below fraction size cut point
viii. Calculate the total particle size distribution
In Figure 52 the breakage process for the CSS50mm simulation can be seen when the chamber
is about to be emptied. It is clearly seen how the particles are broken sequentially in a number
of compression events. Even though a conventional cumulative size distribution curve is not
presented, the picture below gives a visual indication of the product size. It should be noted that
it is a very positive result in itself that the particles do not explode into total disintegration at
the first compression. This was a big problem in the beginning of the modelling and simulation
efforts.
49 |
Chalmers University of Technology | Figure 52 - Snapshot from EDEM showing the breakage process. The fraction particles are hidden and only the
bonds, coloured according to perceived normal force, are shown. When bonds are broken they change colour to
black. The black broken bonds appear in the visualization as long as the two bonded particles are still within 10mm
from each other.
In Figure 53 two pictures are shown at the same time-step. In the picture to the left the meta-
particles are visually represented by the bonds and on the right the fraction particles are
displayed. Both ways of displaying the breakage process are interesting and have their own
advantage. When displaying only the bonds it is easier to see the actual breakage events and the
new progeny clusters that form post breakage. When displaying particles the free fractions
broken apart from the clusters are also seen. Due to the large amount of particles it is difficult
to see the actual breakage however it gives a better understanding of the particle flow in the
chamber. Also, when displaying the bonds it looks like the bulk density in the chamber is
reduced when travelling downwards through the chamber. It can clearly be seen that this is not
the case when displaying the fractions. Of course the reason for this is that more bonds are
broken the further down the chamber resulting in more free fractions in between the unbroken
clusters.
50 |
Chalmers University of Technology | 7. RESULTS & DISCUSSION
The aim of this section is to:
- Make a comparative study between the experiments and simulations
- Discuss the level of compliance between the two domains
7.1 Hydrostatic Pressure
A comparison of the hydrostatic pressure between simulation and experiments can be seen in
Figure 54. The experimental data displays four seconds of operation randomly chosen from the
data set. The simulation data corresponds to the accumulated z-component forces acting on the
mantle for every time position. The total vertical force is divided by the bearing area in order to
calculate the pressure. Also, the mass of the main shaft package has been accounted for by
estimating the masses of all main shaft package components using the CatiaV5 CAD model.
As only a 90 degree section has been modelled the pressure signal from the simulation is
different to the experiment. If modelling the full crusher the DEM pressure signal would have
less variation. With this in mind it is interesting that the results comply as well as they do. Spikes
appear according to the eccentric frequency in both experiment and simulation. This provides
further evidence to the hypothesis that the feeding conditions in the experiments have been
segregated and misaligned.
Pressure CSS50 -Experiment vs. DEM
3,5
3
2,5
]a
P
M 2
[
e
ru
s s1,5
e
rP
1
0,5
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
Time [S]
DEM - CSS50 - Pressure [MPa] EXP - CSS50 - Pressure [MPa]
Figure 54 – Hydrostatic pressure comparison between simulation and experiment for CSS 50mm.
7.2 Throughput Capacity
A comparison of the throughput capacity between simulation and experiment can be seen in
Figure 55. The experimental data is directly taken from Figure 29. In the case of the simulation
a few things needs to be accounted for such as the sectioning and BPM bulk density. Hence the
expression in Eq. 20 is used for calculating the total throughput capacity.
52 |
Chalmers University of Technology | ̇
̇
Eq. 20
Where ̇ is the mass flow rate in the discharge region, is the sectioning factor, is the
bulk density of the BPM clusters and is a factor that accounts for the feed size distribution cut
off point i.e. the amount of small particles that are not modelled in the feed.
The results show a good conformation between simulation and experimental results. This was
expected for the CSS50 simulation due to the higher amount of feed material and more stable
operation. However it is surprising that the CSS34 data corresponds as well as it does.
Throughput capacity - DEM vs. Experiment
500
450
]h400
/n
o350
t[
y CSS34 - DEM
tic300
a p CSS34 - EXP
a250
C
tu200
CSS50 - DEM
p
h CSS50 - EXP
g
u150
o
rh
T100
50
0
Figure 55 – Throughput capacity comparison between simulations and experiments for CSS 34 mm and CSS 50 mm.
The error bars for the DEM columns represent the standard deviation of the mass flow rate data. The error bars for
the experimental columns represent a 5% deviation (subjectively estimated) due to e.g. belt speed deviation or
sample processing error.
When measuring crusher capacity it is normally important to consider how much of a specific
product size that is being produced. This is called the production yield. In the aggregate industry
it is often more important to produce as much as possible of the currently most lucrative
product than to maximize total throughput capacity. As the method for calculating the product
particle size distribution is not developed yet it is currently not possible to obtain the product
yield data from the simulations. When this is available it will potentially be possible to optimize
liner design as well as the operational setting towards specific production targets. More
importantly it may be possible to, from observing the simulations, understand that breakage and
flow behaviour leads to the wanted targets. This kind of understanding will in turn enable
increased accuracy of the analytical mechanistic cone crusher models.
7.3 Power draw
In Figure 56 the power draw is shown for both experiment and simulation at CSS 50mm. As can
be seen the peaks are much higher in the simulation case than the experiment case.
Coincidently the average power draw matches fairly well in the graph however the quarter
sectioning is not accounted for hence the DEM simulations over predict with a factor of four.
From the DEM simulations the power draw is obtained by calculating the total torque on the
mantle due to the tangential forces exerted by the rock particles on the mantle. This method of
calculating the power draw needs to be overlooked as it seems not to give a good indication of
the power draw. Currently the estimation of the total torque upon the mantle is calculated and
exported by EDEM. An alternative approach would be to export every individual rock particle-
mantle force vector and perform the entire analysis in MATLAB instead. This would rule out
53 |
Chalmers University of Technology | 8. CONCLUSIONS
The aim of this section is to:
- Draw conclusion regarding the validation of the developed DEM crusher model
- Answer the previously stated research questions
This master thesis work has generated results and outcome in different knowledge domains.
First of all a virtual crushing environment has been developed and can now be used as a novel
simulation tool. Secondly the performed experiments and simulations contribute to the general
understanding regarding cone crusher performance.
The project was conducted over a longer time period due to several reasons. However, most of
the initially stated tasks and goals of the project have been performed even though some were
relatively difficult to achieve.
Since only one quality simulation has been performed it is not possible to draw conclusions
regarding the validation apart from this specific case. The CSS 50 mm simulation looks very
promising in terms of flow behaviour and breakage process visually. The other CSS setting
needs to be simulated in the same setup in order to draw full conclusions regarding validation.
In order to evaluate if the objectives stated in the early phases have been achieved each
research question will be commented on below;
RQ1. How should a modelling and simulation environment be designed in order to
effectively evaluate performance and design of cone crushers?
When using the virtual crushing environment for design evaluation it is preferable to have as
low setup time as possible. If the purpose of the simulation is to evaluate e.g. a new liner design
it is possible to just replace the old geometry with the new and reuse all other settings. It would
be possible to put in an initial effort to make base models of e.g. all of the crusher models within
the product portfolio. The time required to setup a specific simulation is then substantially
reduced. Of course one would think that a fully automated simulator would be preferable.
However, due to the complexity of the model it is a potential risk if all of the details and
limitations of the simulator are not fully understood.
It would be preferable to predefine a large number of meta-particle models to choose from in
order to enable an easy assembly of the rock population. When a material model has been
developed and calibrated it can of course be reused for later projects. A library of calibrated
BPM models would be a very good feature.
If performing crusher simulations on a regular basis the post processing needs to be more
automated. Currently the data is exported manually to MATLAB or MS Excel where a lot of
manual work is done in order to convert the data to valuable information.
RQ2. To what extent is it possible to use DEM for predicting crusher performance?
The presented results suggest that the performance can be predicted in a good way when it
comes to mass flow and pressure. More work has to be performed on the post-processing in
order to be able to draw conclusions regarding size distribution and power draw.
RQ3. How should a DEM simulation properly be calibrated in order to comply with real
behaviour?
In this work single particle breakage tests have been successfully used as the basis for
calibration. The approach of using SPB and not uniaxial strength test or Brazilian test gave a lot
of insight regarding the influence of compression mode. The standardized breakage tests
generate more statistically predictable data however they do not really capture the complex
55 |
Chalmers University of Technology | breakage of irregular particles with flaws and sharp corners. In future work interparticle piston
and die tests will be evaluated as a suitable method for calibration. Potentially the best result is
gained if conducting both uniaxial strength tests as well as SPB and IPB tests.
The simulation approaches for finding corresponding settings that give good breakage
behaviour have been semi-systematic in this work. Based on the experience from this work a
more statistical approach should be applied. By using the statistical method design of
experiments (DoE) for choosing settings and utilizing batch simulation mode, the correct
strength parameters can be found more efficiently. It would be interesting to develop a
calibration procedure where the data from one simulation is processed and used in an
optimization algorithm for choosing the next set of values.
RQ4. What level of accuracy can be reached by using DEM for simulation of existing cone
crushers?
This question is difficult to answer however from the work it can be concluded that it is all
about trade-offs. It is possible to model and simulate a small section of a crusher very accurately
in terms of capturing most of the size distribution. However, the material model quality suffers
greatly if the full crusher should be modelled. Currently more computational power is needed in
order to simulate a full continuous cone crusher operation with a good material model.
The computational time for the CSS50 simulation was around 400 hours and the number of
particles was 92500. This is enough for filling the chamber to a fairly good choke fed condition
momentarily. Hence if feeding and discharging continuously, only slightly more particles are
needed in order to obtain steady state condition. This leads to the conclusion that in order to
model the full crusher in steady state we need to use 300-400K particles which would give a
simulation time of around 1365 hours or 57 days, which is not feasible. Simulating a full crusher
using this methodology hence requires a computational capacity scale up by at least a factor of
3-6.
RQ5. How does a change in close side setting influence the internal particle dynamics and
crushing operation in the crushing chamber?
If considering both results from experiments and simulations, there is an indication that when
increasing the close side setting the,
throughput capacity increases
power draw decreases linearly
hydrostatic pressure decreases linearly
particle size distribution gets coarser
reduction ratio decreases
56 |
Chalmers University of Technology | 9. FUTURE WORK
The aim of this section is to:
- make recommendations to future work in order to bridge the problems observed in the
project
- make recommendations on what to focus on in the future development of DEM crusher
models
Even though a substantial set of parameters and conditions have been captured in the DEM
model there are a number of critical aspects that need further improvement;
1. Batch feed - Due to computational intensity restrictions only a relatively small batch size
has been simulated. In order to be able to model more material a higher computational
capacity is needed. Also a slightly different approach is needed for introducing the meta-
particles in a continuous manner.
2. Sectioning - Due to computational economy only a 90° section of the crusher has been
modelled. Of course it would be preferable to model the full crusher.
3. Packing density – Due to the packing density in the BPM there is a problem with mass
conservation.
During the late phases of the project a new meta-particle replacement method was developed.
It was not possible to test it in this work however this will solve the issue of not being able to use
the full rock population that was developed.
In this work the built in bonded particle model in EDEM has been used. It is also possible to
write a customized mode by altering source code published by EDEM. This will enable better
control of the strength distribution and other BPM aspects.
Efforts need to be put into developing the post-processing in order to capture product particle
size distribution and power draw in a better way.
57 |
Chalmers University of Technology | Abstract
This master´s thesis has been conducted in cooperation with Atlas Copco Compressor
Technique in Sweden. Atlas Copco Compressor Technique is a large incumbent within the
compressor market and the company is investigating how the organization can adapt to
digitalization. The purpose of the thesis is to study how the focal company can change its
business model to facilitate value creation and increase efficiency, given digitalization.
During this thesis, a literature study of industrial digitalization and technologies served both
as basis for the analysis and also as background information to gain an understanding of the
situation of Atlas Copco Compressor Technique. To ensure an offer that is appealing to
customers, the analysis was also based on empirical findings from case studies of the focal
firm and nine customers within five different industries to understand the customer needs. The
study was focused on the current situation and the internal operations, while the purpose of
the customer study was to identify current and future needs, application of the products and
future development. The analysis was primarily based on the Business Model Canvas
framework developed by Osterwalder and Pigneur (2010) which consists of eight building
blocks that represent the business model of a company. Each building block was studied
separately with regards to the information found on industrial digitalization and the case
studies in order to obtain a structured approach on the analysis.
From the analysis, three business opportunities were derived which are presented in this
report as three Business Model Canvases. These opportunities were considered, when
combined, to facilitate a sustainable business. The first business model creates value for the
customer by offering energy efficiency, which was found to be an increasingly important
factor for customers buying compressors in the future. The same business model increases
efficiency for the focal firm in the sense that the products are easily upgraded wirelessly. The
second business model is directed towards customers that already owns a compressor and is
thus preferably used following to the first business model. The second business model is
characterized by efficient maintenance of the tangible products and the value for the
customers is derived from the efficient service operations and increased dependability of the
products. The last business model is developed to be used in situations when the customers
are in need of a new machine. The main value to be gained for the focal firm is to find
opportunities for upsell and to gain deeper customer relations, internal efficiency will be
increased by automating the back office operations related to sales. By allowing software to
analyze real time data of the condition of customers’ machines, Atlas Copco Compressor
Technique can predict the life of their products and approach the customers with new
offerings at the right time.
In order for Atlas Copco to adopt these three business models containing digital elements, a
five step action plan is proposed for the adoption process. It has been concluded that the
change must be driven from within the company, the organization must therefore facilitate
two way communication, provide leadership for digitization and create a shared vision
throughout the organization. In addition to these business models, an alternative customer
segmentation was proposed. For all three business models, it is in this report proposed that the
customers will be targeted according to their digital potential and adoption rate. The
customers with high adoption rate are preferably targeted first when introducing new
innovative products, followed by the other adopters as they have been influenced by the early
adopters.
Keywords: digitalization, incumbent firm, business opportunities, value propositions, customer segmentation |
Chalmers University of Technology | 1. Introduction
This section will introduce the background, problem formulation, purpose of the thesis and
limitations of the thesis. Furthermore, a disposition of the thesis will be presented.
1.1 Background
Automation and digitization is transforming the modern economy (Weinelt, 2016). Today, we
live in a digitalized age where technologies such as social medias, analytics tools and mobile
phones are predominating the market (Westerman et al., 2012). In the year of 2016, there
were over 2.6 billion smartphones users around the world (Piejko, 2016). The same year,
internet usage reached a penetration rate of 50,1% which is approximately 3,7 billion internet
users around the world (Internet World Stats, 2017). The spread of internet into people’s
homes and mobile phones has given companies the opportunity to reach their customers
around the clock. (Wessel et al., 2016).
A concept that has arisen along with digitalization is the Industrialist Dilemma (Wessel et al.
2016). The powerhouses of the 20th century have had a difficult time to adapt to the new
data-driven world (ibid). Incumbent firms’ responses to change are often rigid and tedious
which victimizes them when new technologies disrupt their industries (Bradley & O’Toole,
2016). By the end of this year, 20 % of today’s market leaders will have lost their leading
position to the 21st century corporations due their leading edge of working with new
technologies (Nordström, 2015). Not even in a dominant market position, adapting and
transforming to the digital age is inevitable (Hendersen, 2016). Doz and Kosonen (2010)
argue that it is especially important for companies that have had success for a long period of
time to transform their way of doing things and adapt their business model to the competitive
situation. Otherwise, there is a possible risk of failing (Doz & Kosonen, 2010). The way of
working that used to be right does not necessarily have to be right today. (ibid) Firms must
innovate continuously and embrace new digital technology in order to stay competitive
(Baller et al., 2016). In a study made by Massachusetts University of Technology together
with Capgemini Consulting indicates that 81% of the respondents have seen a substantial
impact in both performance and competitive advantage, mainly, due to digital transformation
has been recognized (Fitzgerald et al., 2013).
Embracing digital technology should involve transforming every part of a company from
marketing channels to a company’s value proposition of what kind of products and services to
be offered (Hendersen, 2016). A transformation involves investing in technology, R&D and
human capital, all at once (Hendersen, 2016). There has been a shift from transforming how
we do technology to transforming the way we do business and altering the entire business
model (King, 2013). As Amit and Zott (2012) argue, a new innovative business model does
not only allow improvements in revenue, costs and risks, it can also create new opportunities
on existing markets as well as new markets.
1 |
Chalmers University of Technology | 1.2 Problem formulation
Compressed air is used both in industry and in the homes of private consumers because of the
wide range of applications (Air Compressor Guide, 2016a). Compressed air is often taken for
granted, both in industry and in our everyday life and few reflect on where the air is derived
from (Air Compressor Guide, 2016). As for the private consumers, compressors are found in
hair dryers, cars, radios and loudspeakers (Egy, 2016). In the industry, compressed air is used
for instance as an energy carrier where only 7,5% of the energy can be used for the intended
purpose (Air Compressor Guide, 2016b). Furthermore, 75% of the cost of owning a
compressor is related to energy consumption (Nanda, 2017). Beyond the use as an energy
carrier, compressors are also used in power tools and cylinders for cooling, cleaning and
transportation of material among others (BIAB tryckluft, 2011).
In many industries, digitization has become a natural strategic move, as when digital
platforms began to replace paper magazines (World Economic Forum, 2016). In other
industries, as for manufacturing companies, the digital transformation is less obvious
(Westerman et al. 2014). Atlas Copco Compressor Technique, hereafter referred to as ACCT,
is one out of four business areas within Atlas Copco Group. ACCT has come to realize that
the digital transformation has already hit the compressor industry and that the only way of
surviving the change is to conform. Studies show that the majority of the actors within the
compressor industry are highly aware of the fact that digital technology can be very beneficial
for their companies, both in terms of products and services in order to enhance the company’s
performance (Schulte, 2016). The question many companies are asking themselves, including
ACCT, is what they should do and how to do it? A study made by Deloitte (2015) shows that
companies in many industries have difficulties with prioritizing digital initiatives and know
where to begin. The focal company’s problem lies in the lack of knowledge and ACCT does
not know how to respond to this prevailing rapid change.
ACCT has experienced a rapid digitization of their high-end customers. Their production
facilities are becoming more digitally adapted and ACCT knows that if they, as the supplier of
compressed air, are not able to offer what the customer is demanding, someone else on the
market will. ACCT is unsure how their value proposition will be affected by digitalization and
how digital technology can affect their business model and their way of working. Research
shows that a business model needs to be changed overtime to create sustainable value in the
company (Achtenhagen et al., 2013). This is often forgotten, especially by companies that
have been successful for a long time, like ACCT. Such companies tend to forget the
importance of adapting to market demand and competitive situation (Achtenhagen et
al., 2013).
In the case of ACCT, the problem can be seen as twofold. Today, they compete with
innovative and technically superior products and services. Nevertheless, actors are starting to
catch up using digital technology. ACCT does not know how to respond to this rapid digital
change and they are unaware of how the digital change will affect their business model and
their way of working. So in order to maintain its position and competitiveness, ACCT must
adapt to the new and forthcoming conditions to avoid being ousted by new entrants or
substitutes.
2 |
Chalmers University of Technology | 1.3 Purpose
The purpose of this thesis is to analyze how Atlas Copco Compressor Technique can innovate
its business model in order to increase efficiency and profitability, given digitalization.
In fulfilling the purpose, the main intention of this thesis is to provide recommendations on
business model content that will create added value for the customer and increase the
competitiveness of Atlas Copco’s business unit Compressor Technique in Sweden within the
coming five years. In doing so, the following questions will be asked:
How could the Swedish air-compressor industry be affected by digitalization within
the next five years?
What opportunities and threats will digitalization imply for Atlas Copco Compressor
Technique in terms of products, processes and people?
How should Atlas Copco Compressor Technique develop their business model within
five years by employing digitalization in order to expand their business through new
revenue streams?
1.4 Scope and delimitations
The thesis only focus on ACCT’s offer and the scope of the thesis is to study the Swedish
market of compressor technology. All empirical data is gathered from actors that operates on
the Swedish market and all conclusions are based on this data. Thereby, the study is only
applicable for the Swedish market. While offices in other countries can be compared to the
Swedish, it is important to be aware of the fact that there are differences in terms of culture
and regulations. Due to the fact that the study only apply for one case, the generalizability for
other companies is rather low.
The process of business model innovation consists of many phases. This research is limited to
the early phase of business model innovation which means that the latter phases such as
change management and implementation methods will not be subjected in detail.
The interviews were conducted with only large successful customers on their respective
markets to get a representative picture of the surrounding environment of ACCT. This
selection is not assumed to affect the outcome due to the fact that the majority of the external
actors that influences ACCT are precisely large successful companies. The study only
concerned the relationship between ACCT and their customers. Relationships in the previous
steps of supply chain was not studied.
1.5 Disposition of thesis
Chapter 2 presents a description of industrial digitalization, related concepts and technologies
and its business implications. This section provides the basic knowledge needed for further
understanding of the thesis.
Chapter 3 presents the theoretical models that will be used in the analysis. The chapter
presents the concept of a business model, the Business Model Canvas and the theory of
diffusion of innovation followed by a more specific description of digital adoption.
3 |
Chalmers University of Technology | 2. Industrial digitalization
This section provides an overview of industrial digitalization and the aspects of the subject
that was touched upon during this thesis. The section presents the concept of Industrie 4.0,
internet of things and simulation to shed light on the meaning of these concepts. In addition,
the environmental perspective on industrial digitalization will be discussed. Digital
transformation is touched upon in a broad sense before the section is finalized with a
presentation of three companies that have adapted successfully to digital technology.
2.1 Industrie 4.0
German industrial companies have been seen as industrial leaders for decades (Kautzsch,
2016). Today, they are also in the forefront of digitalizing the manufacturing industry (ibid).
The German government presented their industrial development strategy 2011 called Industrie
4.0, a name that refers to the fourth industrial revolution (Germany Trade & Invest, 2017).
The purpose of Industrie 4.0 is to make Germany a leading market for advanced
manufacturing by adapting the industrial sector to the new global trends of connectivity and
digitalization (Germany Trade & Invest, 2017). The German chancellor Angela Merkel
stresses the need for the European Union to develop at a higher pace in order to maintain
competitiveness against growing economies in Asia and USA (Bundesregierung, 2015).
Industrie 4.0 is said to include the following: internet of things, big data analysis, cyber-
physical systems (enabled by sensors etc.) and finally, an infrastructure for digital
communication (Carreiro, 2015). Fettke et al. (2014) also add a few concepts not directly
related to production but still associated with Industrie 4.0: personalized distribution and
procurement, systems for personalized product and service development and corporate social
responsibility. The new generation of manufacturing which Industrie 4.0 has resulted in will
increase efficiency in every process, from R&D and product planning to supply chain
management and pricing (ibid). Kautzsch (2016) means that German manufacturing
companies are currently rewriting the rules by using big data to forecast demand, using
various analytic tools to analyze supplier information, customer behavior and competitor
information and developing the use of 3D printing to improve maintenance and equipment
performance of the factories.
Marr (2016) wrote an article in Forbes magazine regarding Industrie 4.0, saying that a firm
needs four components in order to call itself a Industrie 4.0 system. These components are
interoperability, information transparency, technical assistance and decentralized decision
making.
Interoperability: Communication between people, machines and apparatus.
Information transparency: A digital copy of the real system is created to process all
information.
Technical assistance: The ability for the technical parts of the system to support
people in both decision making and when performing physical operations.
Decentralized decision making: Needed to support the other three components and is
originated in organizational culture.
According to Baur and Wee (2015), Industrie 4.0 has been made possible because of four so
called disruptions: the rapid increase in data volume due to higher computational power (the
development of communication and networks are also included in this category), business
5 |
Chalmers University of Technology | intelligence and analytics, new forms of human machine interactions and the increased ability
to transfer digital information to the physical world.
The drivers for Industrie 4.0 are market forces both in the form of application pull and
technology push forces (Fettke et al., 2014). Firms must find their unique way of competing
for the customers in a rapidly changing technological environment (ibid). Fettke et al. (2014)
especially highlights five market pull and three technology push trends that fuels the need for
firms to adapt to Industrie 4.0:
Market pull
Decreased time to market: The speed that new products and technologies are
introduced on the market are increasing continuously, rendering old technology
obsolete.
Batch size one: The demand for personalized offerings increase and firms thus need to
produce smaller and personalized batches.
Flexibility: Flexibility in both product development and production is needed to
continuously deliver unique and new innovative products for each customer.
Decentralization: Is needed to be able to make decisions rapidly in order to maintain
flexibility and speed. Fettke et al. (2014) recommends flat organizations.
Resource efficiency: Because of the anticipated rise in raw material prices and
increased focus on sustainability, resource efficiency is needed to cut costs of
producing products and for goodwill purposes.
Technology push
Mechanization and automation: To support physical labour and increase efficiency in
production.
Digitalization and networking: To support control and analysis of the production.
Miniaturization: Refers to computer hardware needed to store and process
information. New improved technology has enabled storage of large amount of data at
minimal physical space.
Baur and Wee (2015) mean that firms adapting to Industrie 4.0 should consider three actions
starting with gathering data. Value will be created when the gathered data is used in a proper
way, which has proven to be a common challenge for adapters (Bar-Joseph, 2013). To guide
firms in allocating resources to a suitable development program for Industrie 4.0, Baur and
Wee (2015) have developed a “digital compass”. The “digital compass” shows what value is
related to Industrie 4.0 levers, see table 2.1.
6 |
Chalmers University of Technology | Value drivers Industry 4.0 livers
Asset utilization Routing flexibility
Machine flexibility
Remote monitoring and control
Predictive maintenance
Augmented reality for MRO
Labor Human-robot collaboration
Remote monitoring and control
Digital performance management
Automation of knowledge work
Inventories In situation 3D printing
Real time supply chain optimization
Batch size
Quality Statistical process control
Advanced process control
Digital quality management
Supply/Demand Data-driven demand prediction
Data-driven design to value
Time to market Customer co creation/open innovation
Concurrent engineering
Rapid experimentation and simulation
Service/Aftersales Predictive maintenance
Remote maintenance
Virtually guided self service
Resources/process Smart energy consumption
Intelligent lots
Real time yield optimization
Table 2.1. Industry 4.0 levers (Baur & Wee, 2015)
The second action according to Baur and Wee (2015) relates Industrie 4.0 to the long term
strategy in order to capture the full potential of the transformation. Industrie 4.0 unlocks
possibilities for new business models, which has to be noted by manufacturers in order to
understand their competitive environment (Baur & Wee, 2015). Baur and Wee (2015) list the
following examples:
Platforms: Since customers require increasingly customized products, manufacturers
need to gain a profound understanding of the overall customer need in order to offer
customers complete solutions instead of only a product (Geissbauer et al., 2015). In
order to gain such information manufacturers must “own” the relation with the end
customer, who pulls the demand, according to Geissbauer et al. (2015). The platforms
link all actors in the value chain to the customer, providing all actors with that
“ownership” (ibid). Geissbauer et al. (2015) also state that first movers gain a
7 |
Chalmers University of Technology | competitive advantage by establishing such platforms, especially if the platform
enables them to gain a closer end customer relation.
Pay by use: Enables manufacturers to lower the barriers for new customers by
relieving the customers from investments in equipment and total ownership of
products (Tukker, 2004). Since the customer demand is automatically monitored a
close relation between customer and supplier is implied, enabling the supplier to create
customized offers to customers (ibid). Manufacturing firms provide services and
products combined in order to create multiple revenue streams and increase profit,
since offering products alone is a case of diminishing returns (Bates et al., 2003).
Services can also be a mean of differentiation towards competitors and thus avoid
price competition (ibid).
Licensing intellectual property: In cases where there is a prevailing knowledge
asymmetry between the customer and manufacturer regarding the product,
consultancy services can be offered (Baur & Wee, 2015). Consultancy services
implies that the supplier provides the customer with guidance of how to use the
product efficiently in a broad sense (Tukker, 2004). Like pay by use offers licensing
intellectual property is a way for manufacturers to create additional revenue streams,
but by creating value utilizing in house knowledge (Bates et al., 2003; Tukker,
2004).
Monetizing data: There are three ways of monetizing data according to Ross and
Wixom (2017) which are improving internal processes, adding information to products
and selling data. Using data to facilitate rational decision making and improving
processes has proven to be an efficient way of increasing profitability (Ross &
Wixom, 2017). By using data and analytics tools manufacturers can provide customers
with additional information to the product which implies an additional revenue stream,
but such offerings also implies risks (ibid). If the data is not managed and controlled
properly, quality deficiencies may be detected by customers and the manufacturer may
lose its credibility as supplier (ibid). Selling data is considered by Ross and Wixom
(2017) to be the most difficult way of monetizing data, mainly because it requires
firms to adopt completely new business models.
The third action to be considered when adapting to Industrie 4.0 described by Baur and Wee
(2015) is preparation. Manufacturers must prepare in order to facilitate good conditions for
the changes, implied by the transformation, and these preparations mainly includes recruiting
“digital talents” and planning for restructuring the organization (Baur & Wee, 2015). The
most critical obstacles that firms encounter is how to manage the data and cyber security
according to Baur and Wee (2015).
2.2 Internet of things (IoT)
The concept internet of things (here by referred to as IoT) refers to the connectivity of
manufacturing equipment and products among others and is enabled by systems of systems,
wireless communication and big data analysis which will be further explained separately
below (Bossen & Ingemansson, 2016). The IoT systems enable equipment to interpret and
communicate digitally with its surroundings (Internet of things Sverige, 2017). Notable for
such systems is that they should be implemented with care to ensure integrity and safety
(Bossen & Ingemansson, 2016).
8 |
Chalmers University of Technology | 2.2.1 Systems of systems
The concept systems of systems is many independent smaller systems that are interconnected
in a larger, holistic system which enables interoperability and thus a more efficient use of all
systems and a more flexible system as a whole (Bossen & Ingemansson, 2016). The systems
together perform complex tasks by communicating with each other (Rouse, 2012). The
communication is enabled by suitable interfaces between the smaller subsystems (ibid). By
the use of a system controller, sub optimization of the systems are avoided (ibid). In order to
create a proper system controller, knowledge of how each subsystem works, how they
interoperate over time and how the larger entity works and evolves has to be gained (ibid).
The subsystems are often characterized by different fabricate, different lifecycles and
different owners (Bossen & Ingemansson, 2016).
Systems of systems is a prerequisite for initiatives such as Industrie 4.0 and IoT (Bossen &
Ingemansson, 2016). Wee et al. (2015) state in their report that in order to create a robust flow
of information, common standards is the key to handle data from various sources. To ensure
flexibility Wee et al. (2015) stress the need for transparency and sharing information
throughout the whole value chain, both within the firm and outside with suppliers and
customers.
2.2.2 Wireless communication
Wireless connection has enabled new innovative business models for firms, providing
streaming and navigation services, social media, taxi services etc. (Bossen & Ingemansson,
2016). In the movement towards Internet of things, wireless communication must provide
units with low energy consumption, low unit prices and long battery time and at the same time
support the increased amount of data (ibid). In industry, wireless communication is used for
connecting sensors added to manufacturing equipment and for security reasons (Shen, Wang
& Zhan, 2013). One reason for using wireless sensors in production is the need for reduced
amount of cables on moving equipment, so that the equipment can move with higher degrees
of freedom (Bossen & Ingemansson, 2016). Wireless communication may also impact overall
energy savings by providing synchronized production in central plants and smaller production
facilities in the periphery, to optimize production and distribution (ibid).
2.2.3 Big data analysis
Big data analysis refers to the ability to handle extremely large amounts of data of various
quality (Bossen & Ingemansson, 2016). By developing advanced algorithms, patterns,
correlations, trends and other kinds of useful information can be extracted from the ledge flow
of information in order to support decision making (Rouse, 2017). Big data analysis is made
possible by the combined use of certain technological tools which are described in table 2.2.
9 |
Chalmers University of Technology | Data Management A repeatable process to make data readable by maintaining a quality
standard in the data
Data mining Finds patterns in the incoming data
Hadoop Stores large amounts of data on commodity hardware
In memory Enables immediate analysis and thus insight from the data by
analytics analyzing stored data from the system's memory
Predictive analytics Uses statistical analysis to determine the likelihood of different
outcomes by analyzing historical data
Text mining Uses machine learning and language processing technologies to
analyze written texts
Table 2.2. Big data analysis tools (Davenport & Dyché, 2013).
Big data analysis is not only used for fast and rational decision making, big data analysis has
proven to reduce costs because of the cost advantages provided by the new storage abilities
due to Hadoop and cloud-based analytics (Davenport & Dyché, 2013). Increased value for
customers can also be created by using big data analysis in order to better process customer
information and thus generate a deeper understanding for the customer needs (Davenport &
Dyché, 2013). Big data analysis in manufacturing has proven to raise productivity when used
to analyze and improve supply chains according to Gaitho (2015). Companies like SKF and
Sandvik have adopted big data analysis in order to gather information of their products after
they have been sold in order to for example predict damage due to wear and help their
customers use their products more effectively (Bossen & Ingemansson, 2016). Volvo Cars is
another manufacturer embracing big data analysis techniques, developing technologies to
predict the probability of engine failure (Bossen & Ingemansson, 2016).
There are many areas of application for big data analysis and the most influential ones are
according to LaRiviere et. al (2016) predicted demands, predictive maintenance, improving
pricing and for developing radically new applications. The information below is derived from
LaRiviere et al. (2016) reasoning.
Predicted demand: By using all the above mentioned digital technologies, especially
text mining (see table 2.2), web search data can be analyzed to detect customer
demands and needs. Such information can also be found by studying which customer
segment are visiting certain websites, for example based on location.
Predictive maintenance: The IoT can provide a large stream of real time data which
can be analyzed and interpreted in order to in advance predict machine failure. By
using data mining and predictive analytics, predictions of the life of the machine and
its components can be made. The advantage of making such predictions is that the
dependability of the machine increases, since failure can be prevented in advance
based on the actual condition of the equipment.
Improving pricing: The use of data mining, in combination with the other technologies
mentioned in table 2.2 enables firms to relate price responsiveness and preferences to
algorithmically generated customer segments. Such application of big data analysis
10 |
Chalmers University of Technology | may also be used to match advertisers to customers, which has proven to increase
revenue on online advertising.
Radically new applications: Big data analysis may not only be used to improve existing
processes, it can also disrupt them. By finding patterns in large amounts of data that have not
been accessible before, new conclusions can be made which unlocks new opportunities. For
example, ways of curing diseases and ensure local electricity distribution on demand.
2.3 Simulation
Simulation refers to the ability to create a virtual copy of the real world in order to simulate
different scenarios during for example product development and production planning
(Lindström, 2016). Simulation, or virtual reality (VR) as it is also referred as, is defined by
Onyesolu and Felista (2009) as “A highly interactive, computer-based multimedia
environment in which the user becomes the participant in a computer-generated world”. VR
simulation is a way of using and interact with complex and large amounts of data and is made
possible with today's computing and digital technology (Onyesolu & Felista, 2009).
Using simulation during the product development phase decreases the time to market and
development costs because of the reduced amount of labor and reduced need for physical
testing (Pease & Zistl, 2014; Bossen & Ingemansson, 2016). The earlier in the product
development stage mistakes are made, the less costly they are (Nigel & Slack, 2015) and
according to Reality technologies (2016), the company Gabler has increased their profits by
15% since the introduction of the system. Gabler’s virtual manufacturing factory also allows
for virtual inspection of the equipment sold by experts, something that in real life is difficult
because of the logistics of physically moving people but can be done more easily and
frequently due to the VR technology (Reality technologies, 2016). The purpose of such
activities is to ensure a steady and reliable production for the client, by monitoring and
inspection of the product and its performance (Reality technologies, 2016). VR simulation can
also preferably be used to simulate manufacturing of a product in order to develop an
effective and reliable manufacturing process (Andersson, 2015).
2.4 Technology and environment
The way humanity consumes our planet's resources is not sustainable and the consequences
are severe (Globalportalen, 2017). The greenhouse gas emissions due to the combustion of
fossil fuels are among the largest causes for the climate changes, causing the global middle
temperature to rise and thus affects all living beings on our planet (WWF, 2017). In order to
slow the negative spiral of climate change down, measures have to be taken on a global level
and industry has a key role in the process (Globalportalen, 2017). The industry is responsible
for one third of the emissions due to energy consuming production and transportation (ibid).
Production must become more energy efficient and the use of renewable sources must
increase (WWF, 2017). In September 2015 Agenda 2030 was created at the UN summit in
New York, containing 17 goals for sustainable development (FN, 2017). The meaning with
Agenda 2030 is that environment will no longer be considered separate from development
(FN, 2017). Agenda 2030 constitutes of the global sustainability and environment goals which
applie (FN, 2017). Diplomacy Edu (2017) states that standards are the key to reach the 17
sustainable development goals. For example the ISO 50000 series touch upon energy
efficiency aspects (related to goal 9 regarding industry sustainability, innovation and
11 |
Chalmers University of Technology | infrastructure), but also international standards for trade (goal 17) may contribute to a
streamlined international trading system (Diplomacy Edu, 2017).
The Swedish government aims to become world leading in sustainable development and has
taken measures to alter the society into being modern and ecofriendly (Equalclimate, 2017).
Decreased emissions does not necessarily exclude growth and economic gains according to
Porter and van der Linde (1995). By investing in new technology and energy saving
initiatives, companies can increase profit by increasing efficiency (Porter & van der Linde,
1995). Also, since Sweden is moving towards a fossil fuel free and more energy efficient
society, profitability may even decline if not investing in the previously mentioned changes
(Equalclimate, 2017).
2.5 Digital transformation
In order to become a digitalized actor within the manufacturing industry, the process of
transforming needs to be considered. In the section below the concept digital maturity will be
explained followed by a motivation of why firms should engage in the transformation. In
addition, characteristics of successful digitalized industries will also be presented.
2.5.1 Digital maturity
Digital maturity, or “Digirati” denotes both digital innovative products and the force of
driving transformation and movement forward (Westerman et al., 2012). Recent studies show
that digitally mature companies have higher revenue, profitability and shareholder value than
less-digital companies (Heaslip, 2015). In a study made by Massachusetts Institute of
Technology (MIT) together with Deloitte Consulting, strategy was identified as the key driver
of becoming a digitally mature company (Kane et al., 2015). The study also shows that less
digital companies often focus on individual technologies, instead of focusing of the overall
strategy with its scope and objectives. There are two kinds of dimensions of digital maturity;
digital intensity and transformation management intensity (ibid). Digital intensity dimension
contains the company’s technological operations in terms of customer engagement, internal
processes and business models. Transformation management intensity denote the creation of
the necessary capabilities a leader should have to be able to drive digital transformation (ibid).
For both dimensions, digital innovation and digital culture are equally important when
transforming a company to a digital business (Heaslip, 2015). Heaslip (2015) believes that
any company, startup or large traditional company, can transform.
2.5.2 Digital industries
The degree of digitalization does vary between industries (Gandhi et al., 2016). Studies show
that the leading companies are 13 times more digital than average companies in less
digitalized industries (ibid). Gandhi et al. (2016) underline the digital difference in the
economy by writing that “some parts are playing in an entirely different league”. Over the
past 15 years, company's digital assets have been doubled, not only in terms of pure IT but
also within the physical assets of the company (ibid). There has been an increased usage of
different forms of transactions, customer and supplier interactions and the internal processes
of a business such as activities at back office has become more digitalized (ibid). Gandhi et al.
(2016) distinguish between sectors and their potential for digitalizing their processes, see table
2.3.
12 |
Chalmers University of Technology | Table 2.3. Presentation of different sectors digital potential.
Gandhi et al. (2016) mean that companies can increase their digital ability in many ways, for
example in how they integrate digital tools in their everyday activities but also in the way of
working with customers and suppliers. The research performed by Gandhi et al. (2016) have
identified key attributes of digital leaders that can be used as a framework for other
companies. These attributes have been divided into three categories; digital assets, digital
usage and digital workers (Gandhi et al., 2016).
Digital assets is a measure of how much a company invests in digitizing their assets,
such as hardware, software, data and IT services (Gandhi et al., 2016). It can be smart
buildings, smart vehicles, big data or IOT systems to increase the performance of
equipment and supply chains (ibid)
Digital usage means which extent a company involve their customers digitally
(Gandhi et al., 2016). It may concern digital payments, digital marketing or usage of e-
commerce platforms (ibid). Companies with a high digital usage often manage back
office with different software applications (ibid).
Digital workers involves the extent of which employers use digital tools in their daily
work to increase productivity (Gandhi et al., 2016). Within this category, there is a
13 |
Chalmers University of Technology | huge gap between the best and the inferior in the economy (ibid). The study also
shows that companies priorities digitizing different processes, for instance health care
organizations possess high technological tools while their payment process are very
manual, less than 20 % of the payments are done digitally (ibid).
2.6 Digital technology adopters
In this section, three firms adopting digital technology are reviewed as illustrations of how
firms may use digital technology in a successful way. Siemens is considered to be a
conservative, product based firm that due to a pro-innovation culture are successful in
adopting new technology (Littlefield, 2015). As an illustration of firms using digital
technology to increase efficiency, SAPs application for supporting service technicians is
presented (Oswald & Schreckling, 2015). Finally, a description of how Volvo Cars is using
digital technology to change the performance of their cars is presented.
2.6.1 Siemens – Combining high tech products with services
Siemens has recognized the need for proactive maintenance on industrial products, since their
customers’ production plants are developing into being more complex and refined (Ratkovic,
2013). The firm is incorporating sensor technology and intelligent information processing
technology into their technically advanced products to enable individually tailored service
contracts offers (ibid).
The sensors and digital technology enables Siemens to wirelessly connect to the machines to
gather and process the continuous flow of real time data (Ratkovic, 2013). By harvesting
information from all connected machines worldwide, Siemens create a bank of information
from which the value for the customer is extracted (ibid). In utilizing big data analysis to
process information from all active units, Siemens can predict the life of single parts and thus
offer predictive maintenance services (ibid).
In addition to predictive maintenance services, Siemens is developing an online platform for
spare parts (Ratkovic, 2013). The vision is that by 2025, an application would be developed
that recognizes components material numbers from pictures (ibid). When a predictive
maintenance system discovers that a component is about to break a service technician gets
noticed (ibid). The service technician would then photograph the component and use the
visual recognition application to send the information to the service center who organizes
delivery of the right parts at the right time (ibid).
If a spare part is not available Siemens are hoping to, by using additive manufacturing
technology, decrease lead times on components and thus ensure just in time delivery
(Ratkovic, 2013). Ratkovic’s vision is to install replacing components within 24h after the
customer has recognized a defect (ibid).
To inspect a problem, a service technician would use a device such as glasses or template to
study a machine virtually (Ratkovic, 2013). Augmented reality will in the future enable
Siemens technicians to easily inspect the highlighted problem by having the ability to see
through the machine (ibid). Such systems would also hold information regarding service
history and notes from other technicians who have previously examined the product (ibid).
14 |
Chalmers University of Technology | 2.6.2 SAP – Augmented reality apps and wearable technologies
Customers are using smart gadgets daily but not in workplaces (Powell, 2016). Smart Gadgets
are described by Powell (2016) as smartphones, smart glasses, smart watches and even smart
t-shirts. SAP is developing ways for enterprises to incorporate smart technology and
augmented reality to enhance business performance (ibid). SAP has recognized an increased
demand for not only the use of smartphones but also for other smart gadgets such as glasses,
clocks and even t-shirts (ibid). When developing augmented reality applications SAP has
chosen to focus on smart glasses and mainly the Vuzix M100 glasses that are considered to be
the most rigid ones on the market (ibid).
SAP has developed and announced the release of two augmented reality apps, the “SAP AR
warehouse picker” and the “SAP AR Service technician” (Powell, 2016). In developing
successful augmented reality applications, SAP has recognized that only relevant information
at the time should be displayed in front of the user's actual reality (ibid). The user interaction
is minimal and data is gathered from many different kinds of wirelessly connected sensors
(ibid).
Service technician scenario
A common problem for service technicians recognized by SAP is the lack of information
when arriving to a site (Powell, 2016). It is common that technicians is not able to perform the
job when arriving to a site and thus has to return to the service station before returning with
more information (ibid). Such activities are expensive and if avoided, costs can be cut and the
technician would become more efficient (ibid).
The SAP application would provide the technician with hands free access to instructions and
models of the equipment (Fargel, 2014). Complete with a recording device and voice
recognition for interaction, the technician could record notes to create a digital knowledge
base (ibid). In the case of a technician needing further guidance or support, SAP is
introducing “expert calling” (ibid). The concept implies that support can be brought directly
to the site by calling an expert who can see the situation through a live stream (ibid). The
livestream would be provided by the camera incorporated in the smart glasses (ibid).
2.6.3 Volvo Cars – Polestar optimization and drive modes
Volvo Cars is using digital technology in their products, enabling the company to offer a
service called polestar which is a service for increased performance (Volvo cars, 2014). The
service entails changes in the software that controls the operations without affecting the
hardware (ibid). This creates value for the service customer at minimal effort (ibid). In
addition Volvo Cars is offering cars with various drive modes, enabling the customer to
choose the characteristics of the car's control system depending on situation (Volvo cars,
2015). The drive modes is standardized packages of software (modules), all ensuring a
pleasant driving experience for the customer (ibid). Modular products refers to a product
which is constituted of a number of standardized building blocks (Golfman & Lammers,
2015). The individuality of each product lies in the combination of these standardized
building blocks, which may also be referred to as modules (ibid). Another advantageous
feature of modular products is the ability to rearrange the modules meaning that the character
of the product may be changed over time (ibid). Modular programing can be used in order to
create modularized products (Technopedia, 2017). The modular program has an executable
setup, consisting of one main module upon which auxiliary modules (executable files) are
added which creates value for the customer digitally (ibid).
15 |
Chalmers University of Technology | 3. Frame of reference
This section will present the methods that will be used to interpret the empirical results in
section 5. Initially, the concept ‘business model’ will be presented and thereafter the Business
Model Canvas, theory of diffusion of innovation and digital adoption theories will be
explained.
3.1 Business models
A business model is a method that describes a company’s way of intercept value into the
business (Financial times lexicon, 2017). A business model includes many different aspects
such as advertisement, prices, sources of revenue and product description (ibid). The model
provides a link between internal and external sources and activities which are essential to a
company’s strategy (ibid).
Research has shown that a business model needs to be changed overtime to create sustainable
value in the company (Achtenhagen et al., 2013). This is often forgotten, especially by
companies that have been successful for a long time (ibid). They tend to forget the importance
of adapting to market demand and competitive situation (ibid). A business model is a
framework for making money, something companies need to be reminded of, argue
Achtenhagen et al. (2013).
3.1.1 Business Model Canvas
A Business Model Canvas visualize a company’s way of creating, deliver and capture value
(Osterwalder & Pigneur 2010). The model consists of nine building blocks, see figure 3.1,
which can be further divided into four main areas; customers, offers, infrastructures and the
finances. The information presented below are derived from Osterwalder and Pigneur (2010)
if nothing else is mentioned.
Figure 3.1. The Business Model Canvas by Osterwalder and Pigneur (2010).
16 |
Chalmers University of Technology | Customer segments
The first building block involves the groups of people the business is aiming for. Different
groups of people can create different customer segments, either if they have different need
and requirements or if they ask for different relationships or distribution channels.
Value propositions
The proposed value is the products or services that are requested by the customers. The
products and services that create value for the customer. It can either be a offer that is similar
to the existing offer but with more value added or it could be a offer that can disrupt the
industry. Each customer segment has a corresponding value proposition.
Channels
This building block forms a bridge between customer segments and value propositions and
describes the communication and how the customers want to be reached. The chosen channels
are of importance as it is one of two touch points between the company and its customers.
Customer relationships
This building block describes the relationships that a company would like to establish with the
customers. The customer relationships can either be personal or automated. The form of
relationship can depend on the following; customer acquisition and retention or/and in the
order to increase sales.
Revenue streams
The revenue building block consists of the different streams of cash generated within the
company. Each customer segment has a corresponding stream of revenue. Revenue streams
comes in two forms, either from one-time customers or from continuous payments by a
specific customer segment.
Key resources
This building block describes the assets that are of importance for the company to work. The
key resources are the resources required to be able to create a value proposition to the chosen
customer segments. These resources can either be of physical or intellectual character or be
financial or human factors. The employers of a business are said to be among the most
important resources while they are at the same time also perceived as the most disregarded
resource (Cleverism, 2015).
Key Activities
The key activities denote the activities that are essential for a business model. In line with
above reasoning of key resources, the key activities are required to create a value proposition.
Key activities can either be software development for a software company or supply chain
management for a manufacturing company.
Key Partnerships
The Key partnership building block constitute the network of partners and suppliers that are
required to deliver the value proposed. Osterwalder and Pigneur (2010) distinguishes between
the following four partnerships:
Strategic alliances: cooperation between actors that are not competitors
Coopetition: cooperation between competitors
Joint ventures: actors that creates new businesses
17 |
Chalmers University of Technology | Buyer-supplier: relationship between buyer and supplier ensure delivery
Cost structure
This building block describes the costs a company incur. Every activity that the business
model performs, every partnership it creates, costs occurs. The importance of cost structure
vary among companies. Some companies are more cost driven while others choose a more
value driven cost structure.
3.2 Digital adoption
In this section the theory of diffusion of innovation and theories of how digital technologies
are adopted are presented. In order to adopt digital technologies, it is important to
understand how companies’ are affected by new technologies. It is also important to
understand the difficulties with adopting new technologies and how a company can overcome
these difficulties in order to become a digital master.
3.2.1 Diffusion of innovation
The diffusion of innovation theory describes the social process of the distribution of an
innovation. The most frequently used theory describes the way an innovation is
communicated by individuals which are members of a social system over time (Sahin, 2006).
The spread of an innovation is epidemic and follows a so called s-curve. Illustrated in figure
3.2, the distribution of the innovation in the early stages of its lifecycle are low (Dearing,
2009). The curve also illustrates that the spread increases exponentially until it reaches a
certain point where the penetration of the market has reached a maximum (Rafinejad, 2007).
To what extent an innovation is distributed and accepted among the members of the system
depends on many factors. The spread depends on the uncertainties and obstacles that are
connected to the transformation into the new innovation, the availability to the
communication channels such as mass media, interpersonal communication, and the
composition of the social system (Sahin, 2006).
Figure 3.2. The adoption curve freely interpreted from Dearing (2009)
The individuals in the social system can be categorized into five groups, the innovators, the
early adopters, the early majority, the late majority and the laggards (see figure 3.3.). The
innovators are the ones that bring the innovations into the system from the outside. These
individuals are not respected by the rest of the members in the system, which often is due to a
common unwillingness to accept changes (Sahin, 2006). The early adopters on the other hand
often hold leadership roles which means that other individuals approach to the early adopter
to receive information and advice on the innovation. (Light, 1998) The information spread by
the early adopters are their subjective evaluation of the innovation and they “put their
18 |
Chalmers University of Technology | personal stamp of approval on a new idea by adopting it” (Rogers, 2003). The difference
between the early and the late majority is mainly that the early majority are deliberately
adapting to the innovation when the late majority are led to adapt to the new innovation by
peer pressure or economic aspects. The category that last adapts to the change are the
laggards. The laggards often has a limited interpersonal network which makes them less
knowledgeable about the innovation and thus do not adapt until members of the other
categories has ensured that the innovation works in a satisfying way. The rate of information
about the innovation is an important factor that affect the adaption-time to a new innovation
and Rogers describes the innovation-diffusion process as “an uncertainty reduction process”.
To decrease uncertainties Rogers proposes five attributes to the innovation, relative
advantage, compatibility, complexity and trialability. To which extent the innovation
possesses any of these attributes depends on the personal perception of the individual in the
network.
Figure 3.3. The diffusion of innovation curve, freely interpreted from Rafinejad (2007)
3.2.2 Barriers for digital transformation
A transformation to digitalization involve challenges and problems (Fitzgerald et al., 2013).
There is not one single solution to the problems that arise during a digital transformation,
which is why it is important to be aware of the barriers to transform, mean Fitzgerald et al.
(2013).
The main barrier that was recognized by Fitzgerald et al. (2013) was leadership. In some
cases, a lack of alignment among senior managers can be the cause of the problem (Fitzgerald
et al., 2013). In other cases, the alignment between managers is not shared with the rest of the
organization which appear to be equally as problematic according to Fitzgerald et al. (2013).
The second barrier to transformation is related to the creation of suitable business and
financial cases (ibid). Many companies have, according to Fitzgerald et al. (2013), had clear
struggles with the return on investment of digitization programs due to uncertainties of where
the impact of the transformation is taking place and its quantity (ibid). The third identified
barrier is governance (ibid). The study made by Fitzgerald et al. (2013) show that 40 percent
of the companies participating in the study had no formal governance model in place to run
the transformation properly. It was also revealed that only 26 % percent of the respondents
had formal KPIs to measure the progress of the transformation (ibid). Hence, the lack of
formal governance models and PKI´s constitutes the governance barrier. The last barrier of
transformation is culture (ibid). A transformation of this kind entails change in work practices
which can cause resistance (ibid). The capability aspect relates to the lack of skills that was
experienced by the participants in Fitzgerald et al.´s (2013) study as a major problem in
digital transformation (ibid). This study showed that the barriers to transform a company lies
in governance, measurement and vision. Fitzgerald et al., (2013) argue for the importance of
reviewing the process regularly to understand what prevents the transformation to become
successful.
19 |
Chalmers University of Technology | 3.2.3 Become a digital master
Companies need to move at the speed of the customers, otherwise someone else will, argues
Heaslip (2016). He continues stating that companies often understand why they should move
at the speed of the society, the shortcomings, however often lies in knowing what to do and
how. In order to become a digital master, Gandhi et al. (2016) stresses the importance of
external focus. To be able to increase the external focus, it is important to understand the
external surroundings and how it is digitizing as well as understand the changes in customer's
expectations (Gandhi et al., 2016). It is also important to understand which companies that
will meet these expectations, both from the industry itself but also from other external actors
(ibid). Once a company has an external focus, strategies can be designed in order to intensify
their use of digital tools (Gandhi et al., 2016). In addition, Bilefield (2016) believes that it is
important to transform through the whole organization to make the digital transformation
successful.
Westerman et al. (2014) have identified two critical dimensions when transforming digitally;
digital capabilities and digital leadership. They aim to answer what companies should do in
order to digitally transform and how they should do it. Digital capabilities aims to understand
how the company master its technological aspect (Westerman et al., 2014). To become digital
mature, a company needs to change the way business are done (ibid). Customer relations can
be technological improved as well as internal operations and business models, according to
Westerman et al. (2014). A company can increase efficiency at work and get closer to
customers using new tools as social medias and analytics (Westerman et al., 2014). Digital
leadership is the bridge from technology to the actual transformation (ibid). Research have
shown that bottom up leadership have not given any successful transformations (ibid). Clear
directions, goals and structure from top management is important to master a transformation
(ibid).
In addition to these dimensions, Heaslip (2016) emphasizes the importance of working in
small teams. He argues that even if the team in question risks everything, the total risk for the
company will be reduced. Small teams can experiment, fail, retest and fail again without
disrupting the whole company, says Heaslip (2016).
Libert’s steps
All organizations are different and thereby, there are various ways to transform one. Libert et
al. (2016) argue for a five-step process called pivot, a model that they are using when working
with clients. The steps below are derived from the framework of Libert et al. (2016).
Step 1 - Pinpoint.
The first step in a transformation of a company is to clarify the starting point. By identifying
today’s business model and assets and understand its strengths, weaknesses and long term
habits, a starting point can be clarified.
Step 2 - Inventory.
The second step involves creating a list of the company’s assets, both tangible as plants and
equipment and intangible as skills and intellectual properties.
20 |
Chalmers University of Technology | Step 3- Visualize.
The third step deals with visualization of what the future will bring for the digitalized
network. The aim is to investigate how the network can create value from new investors,
competitors and distributors, both in term of own assets and others’.
Step 4 - Operate.
The fourth step involves doing a pilot where small amounts of time, talent and money are
spent. This step’s aim is to learn and change fast.
Step 5 - Track.
In the last step, new metrics should be reported such as sales and number of active network
partners. Common metrics such as revenue and profitability are also important but is often
perceived as difficult to use in the transformational stage since it usually take time for models
to create value for organizations.
Maxwell’s steps
In a Harvard Business webinar with Maxwell Wessel (2017), the complexity of a digital
transformation was discussed with focus on legacy and the Industrialist Dilemma. During the
session, he presented a five step framework of what a company should do to manage the
process of change successfully. The summarized steps below are based on Wessel’s (2017)
discussion regarding the complexity of digital transformation.
The first step is according to Wessel (2017) also the most important; admit reality. Things that
were success factors last century does not necessarily mean that they will make companies
great this century (Wessel, 2017). The next step in his process is to recall what you do.
Companies often fool themselves to think that customers buy what they sell. Contrariwise, a
customer rarely buys what a company sells. Wessel (2017) says that companies often miss the
point which is to understand what job a customer want to do. “People do not want to buy a
quarter-inch drill. They want a quarter-inch hole!”, quoted by Christensen, Cook and Hall
(2006). The third step is called “north star” and means establishing and agree on long term
goals for the company. This will decrease the pressure and allow short term failures in larger
extent. The fourth step involves decisions regarding further transformation. Wessel (2017)
means that it could be very difficult to persuading all partners to transform. It is important to
realize that different people have different incentives and only some partners or organizations
may be necessary to continue with (Wessel, 2017). VISA performed this step successfully
when they divided its organization and created a part that focused on the future round the
clock (ibid). In the last step, organizations and partners are established and it is now time to
enlarge the pie, i.e. creating new economic opportunities and create value in the ecosystem
(ibid). Wessel (2017) says that a company does not need to own every asset, enlarge the pie
and use companies in the ecosystem.
In addition to the above mentioned processes, Hammer et al. (2016) present four aspects of
what differs a digital master with a non-digital master. The first aspect entail the importance
of focus (ibid). The second aspect presents the use of agile strategy and the importance of
gradually build the transformation to succeed (ibid). As the third aspect, Hammer et al. (2016)
resemble capturing digital opportunities with a team sport. The fourth and last aspect
emphasize the human involvement and Hammer et al. (2016) mean that the real challenge of a
transformation is the transformation of the employees.
21 |
Chalmers University of Technology | 4. Methodology
This section will present the chosen methodology of the thesis which consists of the design of
the research, an outline of the working process and the methodology behind data collection
and data analysis. Furthermore, generalizability, validity and reliability of the study will also
be discussed.
4.1 Research design
Easterby-Smith et al. (2015) believe that a study of qualitative character is useful for
researchers that are trying to understand particular issues within a specific area. Aligned with
that reasoning, this study was performed as a qualitative study due to its definition and
character. A case study research design was selected because of the choice to investigate a
single organization and its real-life context in depth (Easterby-Smith et al. 2015). The
research design aimed to support the process of addressing the research questions, which
implied a consistency in the activities performed during the project (Easterby-Smith et al.
2015).
A case study research design involves large amount of data and some believe that the result of
a case study research is non generalizable and thereby, cannot represent an entire population
(Easterby-Smith et al. 2015). Others discuss whether a case study research is subjective or
objective (Easterby-Smith et al. 2015). To counteract these risks, one should use multiple
sources of evidence in order to triangulate the information (ibid). Easterby-Smith et al. (2015)
also emphasize the importance of having a distinct research design with specified research
questions and decisions how the data should be interpreted before the data collection starts. In
line with that reasoning, triangulation was used throughout the process of the thesis project in
order to increase the strength of the study. Using different perspectives and complementary
information from secondary data increases the validity and generalizability of the study
(Easterby-Smith et al. 2015).
The replicability of this specific thesis will depend on the character of the environment and
the company in question. Although, it can be argued that the analytical framework can be
used in other contexts with companies from similar industries henceforth.
4.2 Research process
Figure 4.1. The research process of the study.
The research process was divided into three stages; a section of data collection where a study
of the topic and an empirical study was performed, an analytical section and a discussion of
alternative solutions, see figure 4.1. During the first part of the research, the data collection,
22 |
Chalmers University of Technology | an iterative literature study was performed where repeating rounds of literature reading was
performed. The initial stage of the study was also served as base for the following stages of
the thesis. In order to gain a better understanding the compressor industry and the underlying
concepts of digitalization, the following topics was studied: Industrie 4.0, IoT, simulation,
digital transformation and digital technology adopters. In addition, environmental aspects
have been relevant throughout the process. Complementary literature concerning theoretical
frameworks applicable for the analysis was also studied in order to gain a better
understanding of the case which this study was focused on and further support the analysis.
The literature was gathered using written sources such as Harvard Business Review and by
participating in a web based seminars at Harvard Business School.
The initial stage also included an empirical study, consisted of primary and secondary data.
Interviews with the focal company was performed as well as interviews with customers. In
addition to the literature review and the empirical study, a theoretical framework was also
created in order to understand how the focal company could digitalize their business model to
stay competitive.
The second step in the research process was of an analytical character. In order to answer the
designated research questions and fulfill the purpose, this step was essential for the study. The
analytical framework was applied on relevant empirical data and background information. A
business model analysis was performed where each building block of the business model was
discussed. In the last step of the process, based on the analysis performed, suggestions for
alternative business models were created and presented. A discussion of each proposition was
performed and the step was finalized with a recommendation of how a successful digital
transformation can be pursued.
4.3 Data collection
The collected data was gathered from both primary and secondary sources and was of both
quantitative and qualitative character. The data has been collected continuously throughout
the research process.
The data collection used both purposive sampling and snowball sampling. Easterby-Smith et
al. (2015) argue that purposive sampling should be used when the researcher has a clear idea
of what kind of data is needed, which goes in line with the purpose of the study. The selected
data was large successful customers to the focal company. In addition to purposive sampling,
snowball sampling was also used during the study. This method is useful when
knowledgeable individuals are difficult to identify (Easterby-Smith et al., 2015). The selection
of the focal company’s employees was performed using snowball sampling, starting with the
supervisor which further recommended suitable employees.
4.3.1 Secondary data collection
Secondary data was collected throughout the research process, both as complementary data to
other data gatherings and also as standalone information. The data was constituted of
information from competitors and customer to the focal company. The purpose of the
secondary data collection was to get a macro perspective of the head organization and the
remaining actors that were interviewed during the study. As preparation for the interviews, the
current company was studied in order to minimize the risk of asking for information that
could be found at the relevant websites or in other written sources. This increased the quality
of the interviews.
23 |
Chalmers University of Technology | 4.3.2 Primary data collection
The interviews were divided into interviews with employees at the focal company and
interviews with customers to the focal company. All interviews were of semi structured
character where questions were predetermined with a flexible nature. Semi structured
interviews provides a flexibility and it allows the interviewers to deviate if that is needed
(Easterby-Smith et al. 2015). A semi structured interview is a successful method when the
interviewer need a range of insights on a specific topic (ibid). The method encourage a two
way communication which in turn can give more in-depth answers where not only the
answers to the questions are given, but also the reasons (ibid).
External interviews with customers
The external interviews was comprised of interviews with customers to ACCT. There were
nine different companies interviewed and in total 12 interviewees. The average length for the
interviews were approximately one hour. The customer respondents were actors from
industries such as Paper mill, Petro chemistry, Heavy automotive, Beverage and
pharmaceutical that use compressor technology from ACCT. The size of the companies
varied, from micro companies to small and middle sized ones. The interviewees were of
different characters where managers at different levels were asked as well as automation and
operating engineers and maintenances administrators. The purpose was to give as a
comprehensive picture of the customer environment as possible. The questions asked aimed to
understand the customer's’ total demand and their use of compressed air. Factors influencing
the choice of supplier were also discussed as well as thoughts and experiences concerning
digitization. The factors used were chosen based on information about the compressor
industry, presented in the literature chapter. All factors were considered to be important for
the customer in the decision making.
The competitive respondents consisted of main actors within compressor industry, namely
ACCT main competitors. The purpose of that information collection was to map what other
actors are doing and how they are doing it in order to identify what the next step should be.
Internal interviews at ACCT
The internal interviews consisted of interviews with four employees at ACCT. The
respondents were a general manager, a vice president of operations, a process manager and a
lead generation manager. These interviews had an average time interval of one and a half
hours. The interviews covered the following topics; vision and mission, various offerings,
customer demands, application areas, business models and digitalization.
4.3.3 Interview setup
All interview questions were sent in advance in order to prepare the interviewee and ensure a
comfortable environment. It can be difficult to obtain trust during an interview so Easterby-
Smith et al. (2015) recommend keeping the interviewees well informed before and during the
interview.
The interviews were primarily held at the interviewees’ offices but Skype was used when
necessary. The choice of locations is in line with Easterby-Smith et al. (2015) reasoning that
locations are essential for the interview outcome and it should be at a location where the
interviewees feel comfortable. The interviews always started with a short presentation of the
thesis and an outline of the purpose and structure with the interview, this to make the
interviewees comfortable in what to expect.
24 |
Chalmers University of Technology | All interviews were recorded in order to avoid misinterpretations. Easterby-Smith et al.
(2015) believe that recorded interviews makes it possible to listen again and avoid
misinterpretations but it is also essential if the interviews are to be transcribed. In addition,
notes were taken during the interviews, primarily to make the interviewee feel appreciated but
also to help the researcher come up with follow up questions. Both researchers participated in
all interviews, this to avoid biased answers.
4.4 Data analysis
The qualitative data that was collected and summarized has been further explained and
interpreted in a qualitative analysis. The analysis began with open ended questions and moved
towards more precise answers, characteristic of a qualitative analysis (Easterby-Smith et al.
2015). The data analysis consisted of circular and nonlinear reasonings and it was iterative
and progressive.
A deductive approach was used (Easterby-Smith et al. 2015). A deductive approach begins by
using various frameworks to group the gathered data and then look for relationships in those
frameworks. This approach is used when the qualitative research is the main design in the
study which is the case in this research (Easterby-Smith et al. 2015). The analysis has been
structured in the following way, see figure 4.2.
Figure 4.2. The process of the analytical work.
The first step in the data analysis was to transcribe the collected data, namely all interviews
performed. To facilitate the analytical work further, the data was structured into internal and
external data and was further divided into several themes that was labelled with today’s
organization, future developments, thoughts around digitalization among others. The next step
of the data analysis was to identify which framework that would be most suitable for the
analysis. The identification started with a discussion of potential closures of the study. By
identifying that a number of business models were desired for the analysis, it became easier to
understand which tools that were needed to deliver the desired result. The framework
consisted of a Business Model Canvas analysis together with the diffusion of innovation
theory that was later applied at the respondent customers. The purpose was to investigate how
the empirical findings, together with the industry and digitalization facts, could be used in
ACCT’s advantage. After the data was organized and the framework was selected, the
empirical results, both external and internal findings, was distributed into the chosen
framework together with the background review of digitalization and industry facts. The
analytical work continued by identifying emerging themes in the framework that were based
on a combination of customer needs and academic findings of the digital surroundings.
Further, the emerging themes were organized so the result could be displayed, that is in a
number of different business models.
25 |
Chalmers University of Technology | 5. Empirical results
The empirical results have been divided into two sections including internal interviews and
external interviews. The internal interviews consist of gathered information from key
employees at ACCT along with information from company-specific material. The empirical
results external to the focal company touches upon customers and competitors to ACCT. The
chosen customers are players that are considered relevant for ACCT’s future development
and work. The competitive overview consists of a brief description of three large and
inspiring compressor actors containing their competitive strategy and digital orientation.
5.1 Internal environment
The internal environment are based solely on performed interviews at ACCT. The following
employees constitute the empirical evidence below:
General Manager
Vice President of Operations
Lead Manager
Business Process Manager
Compressed air is a prime energy source for many industries, to either a smaller or a larger
extent. As a supplier of compressed air, ACCT is no exception and is present in multiple
industries. The business area Compressor Technique delivers a range of products to their
customers, from industrial, gas and process compressors to miscellaneous equipment related
to compressed air. Each business area is also divided into a variety of divisions, see figure 5.1.
Figure 5.1. The structure of the Atlas Copco Group
26 |
Chalmers University of Technology | Atlas Copco is a product driven customer focused company which means that the products are
produced by Atlas Copco in-house. They offer standardized products that is available for the
larger extent of the customer base, thereby the product driven customer focused approach.
ACCT has recognized many business opportunities in the near future on an otherwise mature
industrial market. There are pending investments for compressor technology in large
industries such as paper, food, pharmaceutical and heavy automotive industry. Although, that
does not necessarily mean growth, according to the General Manager (GM). ACCT is
continuously getting opportunities to replace equipment that can be up to 20-25 years old.
ACCT strives to be at the right place at the right time to catch opportunities for further
investments. ACCT’s highest generated value is found in high end customers that owns more
than one compressor.
5.1.1 ACCT´s vision
The vision of ACCT is that each division should be first in mind and first in choice. ACCT
should be the number one brand, according to themselves. The Atlas Copco group’s vision is
yet more complex due to the possession of multi-brands. At that level, the vision is that the
Atlas Copco brand should be number one and the multi brands should be number two. The
strategy of ACCT is to offer top class products. A parable that the GM addresses is the car
industry with Skoda and Audi. He means that some actors compete with price and their
strategy is to be the cheapest actor on the market. ACCT compete within the premium
segment, where top class products is the essential and price comes as secondary.
In order for ACCT to sell more advanced premium products, there is a need towards the
customer to be able to explain the value and the reason why the customer should buy an
ACCT product. Part of ACCT’s strategy is to have a coordinated salesforce, an extensive
aftermarket and a service divisions that can demonstrate that value. Another strategic choice
of ACCT is to develop and produce the majority of the core products in house. The advantage
of developing and producing in house is, according to the GM, that ACCT owns their
intellectual property and there is no risk of information and knowledge leakage. He continues
by saying that there is always a discussion of ownership when it comes to co-development.
Therefore, ACCT does not think they would benefit from co-developments.
“We started earlier. We had a head start, let’s not make them smart.”
- Vice President Operations
What gets measured gets done. The foregoing is a well-used proverb at ACCT. The GM says
that the challenge with measurement is to actually use all that information. He has worked for
Atlas Copco for over 35 years and at the start, there were much less measurements than it is
today. The measuring tools were simply not there, he says. Today, we measure a lot but just
within a few years, I am sure that there will be even more measurements, says the GM.
27 |
Chalmers University of Technology | 5.1.2 Core competences
ACCT presented four core competencies within the area; product development, standardized
production process, distribution and sales and services.
Product development
ACCT is constantly striving to be at the forefront of research and development while
continuously produce innovative products. Their visions is to offer the best solutions to all
customers and to produce products that are most beneficial for the customers. In terms of total
cost of ownership, ACCT also tries to satisfy customers in best way possible. Safety and
energy saving are two factors that always have been are highly prioritized at the company.
Standardized production process
ACCT is said to be good at selling grey boxes. A standardized and good production process is
essential to obtain high class production and cost leadership. A combination of producing
high quality products and having lowest production costs because of a standardized
production process brings health margins in the industry, says GM.
Distribution
The majority of the machines are sold through direct channels, so called direct sales. There
are physical customer centers in 80 out of 190 countries. ACCT puts a lot of effort of sales
and has an extensive network of sales force.
Sales and Service
ACCT has dedicated trained operations for both sales and services. In ACCT, more than 50%
of the employees are related to service.
5.1.3 Current digital transformation
ACCT is in the process of transformation. The Vice President of Operations (VPO) estimates
that only 30-40% of the company is digitalized today. There is a long way to go, he says.
ACCT presents two rules of thumb concerning digitalization, see below.
Two rules of thumb:
Do not digitalize for the sake of it. A digitization initiative should solve a problem. It
is only then a company will benefit of doing something in a digital way.
One should not wait until the final product is complete. It is better with smaller steps
and the product does not necessarily have to be perfect in the beginning.
ACCT considers that at least part of their organization is already digitalized. According to the
VPO, it is fair to say that ACCT is still ahead of competition concerning their connected
machines that they are offering and the website’s offer. Nevertheless, there is a fear that the
competitors will catch up soon, says the GM. He continues by explaining that all of today’s
major actors offer connected solutions, energy saving solutions and extended websites where
the customer can go in and study the behavior of a particular machine. He ends the discussion
by stating that these solutions are becoming more commoditized.
ACCT is working on predicted maintenance solutions where the machines can predict with
exact precision when the machine needs service. ACCT is not where they want to be yet, it is
rather an ongoing process of work today, says the VPO. The website is also in a development
28 |
Chalmers University of Technology | stage where the software is not entirely integrated yet. Today, it is rather standalone. The
vision is to provide one portal where all information can be found, from the machine's’
behavior to technical information and administration tasks such as invoices sent. This implies
less personal interaction that will allow the back office to do more value added activities
instead, says the GM. The purpose is to create one digital consolidated contact point at
ACCT.
The service division has the vision of having every machine above 30 kilowatts connected in
order to push data to an IoT platform. This in order to learn internally and benefit from data
that subsequently will offer the customer transparency in terms of the information they will be
able to see on the IoT platform. An example of a large digital initiative performed by ACCT
is the initiative of going digital at the service department with the service technicians. The
characteristics of the technicians were 40-50 year old men which have not grown up in a
digital world. The initiative of a digital mobile solution took ACCT ten years to finish. The
mobile initiative started 2007 and in March this year, ten years after the start, ACCT finally
rolled out the final mobile phones to the three remaining countries in world. Today, every
service technician knows how to use a mobile phone in the daily work. Now, there is a need
to change the software in the mobile device and the same exercise with the technicians will be
done once again but this time, the initiative will be performed in seven months. A clear
demonstration of how fast the speed of digitalization is, says VPO.
Figure 5.2. How overall changes are performed at ACCT.
The overall changes at ACCT is performed in smaller project with a learning phase, see figure
5.2. It is only after the learning phase they can evaluate and see how the company can mature
further. Finally, the last stage in their process is implementing on larger scale. It is a challenge
to find the perfect package that will fit everyone, says GM. The latest SAP implementation is
one example of that. All of the respondents agreed that difficulties arose both during and after
this particular project due to lack of culture cohesion, leadership, information and knowledge.
“The world hates changes but it is the only thing that has brought progress.”
- General Manager
5.1.4 Business model
A component that has been a part of both yesterday’s business model and today’s is the direct
sales. ACCT has emphasized the importance of having direct contact with the customer to the
greatest extent for years. The interviewees do not think that direct sales management will
change in the future.
The GM says that the content of the business model has changed the last the last ten years.
Yesterday’s business model covered among other things, traditional sales management. The
customer called the sales engineer and asked for a meeting. The sales engineer visited the
customer and had a presentation of the different options of compressors. In short terms, the
sales management was characterized by human interaction. The maintenance work aimed
29 |
Chalmers University of Technology | primarily to help when a compressor broke down. Preventive and corrective maintenance
characterized ACCT’s service management.
Today’s business model is characterized, in addition to what's mentioned above, by a sales
management where the customer uses the website in a larger extent. The customer contacts
the service engineer and explain what they have found on the website and what he/she is
interested in. It has become clear that the customers are more informed today on the ground of
digitalization. The use of online stores has gained importance and ACCT has realized that
they must adapt to that. It has also been realized that there is a market for second hand
machines in addition to the newly produced machines. This is part of today’s business model.
As mentioned above, machines are more compatible with the internet today than it was ten
years ago. Many in-house processes have become more digital such as the channels between
customer and ACCT. For example, the website is used in a larger extent today as a channel of
communication.
5.2 External environment
The external environment of ACCT constitutes a selection of customers and competitors that
are presumed to be representable. This section will present the customer's compressor use of
today and in the future and it will also present their own production development and their
view on digitalization.
The customers which constitute the empirical evidence are presented in table 5.1. Each
interviewee will hereby represent the company for which they are employed and the company
are being named after industry and size. In those cases where two interviews have been
performed at the same company, they will together represent the company in the analysis
below. In addition, companies from the same sector with the same size have been named (1)
and (2).
Sector Size of company Interviewee
Paper mill Large Automation engineer
Strategic purchaser
Paper mill Medium Purchase manager
Petrochemical Medium Procurement coordinator
Procurement specialist
Petrochemical (1) Small Operating engineer
Petrochemical (2) Small Operating engineer
Pharmaceuticals Large Accountable engineer
Beverages Micro Process manager
Heavy automotive (1) Large Maintenance administrator
Heavy Automotive (2) Large Plant maintenance manager
Maintenance engineer
30 |
Chalmers University of Technology | Table 5.1. The respondent customers.
5.2.1 Compressor application
In the following section a description of what kind of equipment that is used to produce
compressed air and how that air is used in different industries are presented.
Paper mill
The two paper mill companies use, regardless of the difference in company size, compressors
in similar ways. Both companies have a number of large machines of which one or two are
frequency controlled 1and a few of them are considered to be very old. The compressors are
grouped together to form a compressor station that provides the production facility with
compressed air. The machines produce a constant but varied flow of air around the clock,
depending on the need for the production. Normally, these compressors are operating on 60-
70% of their total capacity. The air is used to operate control valves for the control systems
and in the packing lines.
The large paper mill company has a service agreement with their compressor provider ACCT,
which is the main provider. The agreement implies that ACCTs service technicians perform
maintenance on the machines based on operating hours which are not wirelessly connected to
ACCT.
Petrochemical
Like the companies operating in the paper mill industry, the interviewees’ firms operating in
this industry use compressors in similar ways, despite the differences in size. All three
companies have a number of large compressors in their compressor station. The small
petrochemical company (1) though, also uses one large specially made machine as part of the
core production. The level of automation and remote control varies between the machines and
none of them are connected to ACCT. All companies have service agreements with both
ACCT and a rival supplier. The service technicians’ share maintenance reports which is used
as basis for ordering spare parts to keep in store in-house preventively.
Pharmaceutical
The large pharmaceutical company use ACCT compressors in 95% of all the cases, some
frequency controlled and others not. They use compressed air to operate controlling valves,
ventilation and for fume hoods. Compressed air is also used as power source in “explosive
areas” certified areas of the production plant. They have had a collaboration with ACCT who
offered a service for energy efficiency. The air losses due to leakages decreased from 34% to
4% after the service was performed.
Beverage
One micro beverage was studied that used two non-frequency controlled compressors. The
second compressor was recently bought from Blocket.se, which is a trading site for used
items. Both compressors are made by ACCT but they only have a service agreement for the
older compressor. The service agreement on the newly purchased used machine is with a third
party not related to ACCT. Compressed air is used as an energy source in every step of the
manufacturing process, both in the brewery and in the tapping activity.
1 Allows for changing the load on the machine, it may also be referred to as variable speed
control (VFDs, 2017)
31 |
Chalmers University of Technology | Heavy automotive
The way compressors are used at the two studied heavy automotive companies do not differ in
general. Both companies are of the same size and use 10-15 compressors at their facilities of
which about half constitutes for their compressor stations where none are frequency
controlled and most are old, which implies expensive renovations.
The production process is organized as a fishbone and is only operational during daytime. A
fishbone process constitutes of a main assembly line to which smaller kitting stations and
even smaller lines are attached along the line, according to the heavy automotive company
(2). The production is dependent on the compressors since they power all “power-tools” used
along the lines.
The heavy automotive company (1) has a service agreement with ACCT, where the technician
gets a sms when it is time for service. To control the machines efficiently, the heavy
automotive company (1) has made their own superior control system for all compressors to
achieve interoperability. The purpose of the control system is to integrate the operations of the
machines to run them more efficiently. Currently, negotiations are pursued with ACCT to
lease a new unit for such system.
5.2.2 Ownership
Three forms of ownerships were discussed during the interviews when interviewees was
asked about what form of ownership they have and prefer; full ownership by the customer,
leasing contracts and buying air per use. All interviewees have self-owned compressors while
one of the respondent also has chosen to lease one machine. Buying air per use was a concept
only discussed in theory since none of the respondents are currently buying air per use.
Both paper mill companies have a policy of funding all investments with their own money.
One respondent at the medium sized petro chemistry company explained that when the
company is performing well, binding capital in equipment is preferred.
Leasing
The small petrochemical company (1) is leasing one of their compressors. The reason for
leasing is the inability to invest in new equipment at the time and thus leasing became a valid
option. The experience of leasing the machine has been positive overall but the interviewee
mentioned difficulties in integrating the control systems to achieve interoperability, due to
limited access to the machines control unit.
The possibility of leasing machines was also brought up during the discussions with other
companies in the petrochemical and paper mill industries. The large paper mill company
concluded that it would be unfavorable for them since costs are difficult to estimate and the
owner of the machine controls the equipment. Even though the interviewee at the company
experiences an increased transparency in the relation between buyer and supplier, having a
supplier controlling part of their machinery is today out of question. The reason for that is,
according to the interviewee, the lack of control over equipment connected to their own
processes.
Buying air per use
During the interviews, buying air was discussed by the small (1) and medium petrochemical
as an alternative ownership. Both companies brought up the option during the discussion
regarding forms of ownership. The small petrochemical company (1) reasoned that the
32 |
Chalmers University of Technology | advantage would be that someone else would take care of the maintenance, otherwise the
interviewee was unsure about the advantages. The interviewee at the company concludes that
buying air per use is not considered possible today but possibilities may appear in the future.
The medium petrochemical company made the connection between their own situation, being
located at the Swedish petrochemical cluster, and large industry complexes in Germany. In
those large complexes, it was said that all companies are provided with air by an external
actor. However, according to the interviewees, buying air per use is not considered favorable
today compared to owning machines due to the company's prevailing economic situation.
However, it may be relevant in the future.
5.2.3 Buyer-supplier relationship
During discussions about the relationship with ACCT as a supplier, our respondents described
the purchase process as being similar to a collaboration characterized by a high level of
transparency. One of the four respondents, the large company operating in the paper mill
industry argued that relations with all suppliers are of high importance. The company is
expecting their suppliers not only to fulfill a need but to engage in improving the overall
performance by using expertise that cannot be found in house. Another respondent, also in the
paper mill industry mentioned that ACCT, by their own initiative had offered a service for
measuring and increasing efficiency on the air compressors. The service was considered to be
a collaboration to help the company improving their efficiency. The pharmaceutical company
stated that, in general, a close relation with suppliers are preferred in order to maintain high
quality on the product by gaining access to service. The interviewee continued arguing that it
is preferred to use one supplier for all machines since they are known and accepted by the
staff working close to them daily.
Interviewees at the medium size petrochemical company and the large paper mill company
stated that they have had relations with ACCT for over 40 years. The interviewees are thus
confident in knowing that ACCT obtain technological competence. Regardless of the long
relationship though, both companies have chosen other suppliers as well during that time
when the price has been favorable. The reason for that is that ACCT's products are expensive,
but when estimating total life cost, the price is competitive according to one interviewee at the
medium sized petrochemical company. This is a statement that is shared by the small sized
petrochemical company (1). A competitor was able to offer a suitable solution compatible
with certain documentation requirements for the medium sized petrochemical company. One
interviewee at the company stated that the same competitor had difficulties in delivering an
attractive price on service, and thus rejected the offer. The reason for the competitors’ high
service prices was according to the interviewee related to low availability and high execution
time.
When describing the perception of ACCT as a company, all interviewees stated that ACCT
has a strong brand and is a dependable Swedish supplier who provides high quality services
and products. The impression that the service technicians are at the right place at the right
time is shared by the large paper mill company, the micro beverage and the medium
petrochemical company. All companies state that ACCT has a technical advantage and a wide
range of products. The overall experience from the small petrochemical (1) company is that
ACCT is putting effort into developing a suitable product and at the same time are open with
their own abilities and limitations.
33 |
Chalmers University of Technology | There are different benefits with different suppliers according to the medium size paper mill
company, stating that some prefer one brand while others another. Even if that is the case
though, the interviewee finds guarantees and installation most important when choosing a
supplier. At the time when the small petrochemical company (1) chose ACCT as supplier for
their newest compressor investment, the product design which enabled easy installation was
one of the determining factors.
5.2.4 The purchase process
A majority of the interviewees stated that many interests have to be taken into account when
buying a large compressor. Both of the heavy automotive companies stated that the purchase
process is initiated with a list of product specifications, such as desired capacity, energy
efficiency and standards that are sent to the supplier before the negotiations begin.
Buying a large machine is by the interviewees considered to be a large investment and
companies are thus executing the purchase process in project form with a team of people with
various expertise. One interviewee at the medium petrochemical company referred to a
compressor purchase as one of the larger purchases they do. The respondent continued stating
that people from technology, operations, maintenance and purchase departments have to be
involved in such purchase process. The interviewee also explained that both site managers
and the Vice President had to be involved in the decision making because of the size of the
investment. Commodity managers are the front figures and the steering group of the project
makes the decisions, based on suggestions from the purchase and technology experts, the
interviewee concludes. At the other two petrochemical companies and the two paper mill
companies interviewed, the purchaser´s role is similar to one another. The purchaser’s role in
those companies is to lay out various financially healthy alternatives and monitor the
discussions to ensure a rational process is maintained. In the case of all interviewed
companies, if the company is part of a larger group large investments are often coordinated on
a central level and for example compressors for the sites are bought at the same time.
5.2.5 Purchase factors – Present classification
During the interviews factors regarding desired characteristics of today’s compressor
offerings were discussed. The factors were price, quality, service, delivery precision, lead
time and energy efficiency, see figure 5.3. The interviewees stated the relative importance of
these factors, as well as motivating their choices and clarified their perception of each factor
and how they are related. Factors classified as highly important are represented with a high
score in figure 5.3.
34 |
Chalmers University of Technology | Figure 5.3. Illustration of the relative importance of the purchase factors.
During the discussion of the importance of price and its impact on the purchase decision, a
lifecycle perspective on investments was discussed with representatives from the large paper
mill company, the medium petrochemical company and both heavy automotive companies.
Both the large paper mill company and the medium petrochemical company stated that
compressors have a long life, which is why the initial cost of purchasing the machine only
represent part of the total cost of buying a compressor. Both continued their reasoning by
saying that both energy efficiency and quality affects the total cost of the investment, and thus
becomes a matter of price. A low quality product might have a shorter life cycle, it may also
imply increased costs for service and maintenance, cost of disturbance in the core production
and higher operations cost according to all interviewees. As shown in figure 5.3, the large
paper mill, medium and small (2) petrochemical companies consider price to be important
since they consider it to be related to quality. A high quality product imply lower costs
throughout the life of the product because of reduced maintenance costs and a longer product
life. The interviewee motivates the choice by stating that the final decisions are based on total
cost analyses.
The interviewee representing the small petrochemical company (1) mentioned that the
negotiations process begins with price but in the end technological superiority exceeds price.
The micro beverage representative agrees with the reasoning and said during the interview
that price is important, but not on the expense of high quality and dependability.
Dependability is described by uptime by the interviewee, who stresses that the production
35 |
Chalmers University of Technology | relies on the compressors. One interviewee at the large paper mill company stated that price
only affects the deal if the price is so high that they cannot afford it.
Even though one of the representatives at the medium size petrochemical company stated that
price was the most important factor, he still expressed the need for high quality. It was also
stated by a representative at the medium petrochemical company that if they have good
experiences with a supplier and know that they will get good service, price and delivery time
are less important. The interviewee continued saying that price is always negotiable.
Interviewees at the small petrochemical company (1), the micro beverage, the medium paper
mill company, the heavy automotive company (2) and the large paper mill company all
considered quality and service as the most important factors. It was stated at the medium
paper mill and pharmaceutical company that quality and service is closely related to
dependability and lifespan of the product. The pharmaceutical company thus consider
dependability to be of highest importance, followed by quality and service which together
constitutes for dependability.
During discussions regarding service, the large paper mill and the micro beverage companies
stated that they considered it to be most important. Interviewees at both companies stress the
importance of dependability of the product performance and relates this to service offerings,
since their production is relying on the compressors inter alia. Service was considered most
important when buying a compressor by many interviewees and good service, as perceived by
all interviewees, is related to availability and it is appreciated that technicians can be in place
at the right time. Delivery precision is not as important for product delivery and spare parts,
since they keep them on stock, according at the large paper mill and medium petrochemical
company. It was also mentioned by the interviewee at the medium sized petrochemical
company that if the delivery precision is good, lead time becomes less important.
In the discussions about energy efficiency, there were different opinions. Energy efficiency
was by the small petrochemical company (2) considered most important while interviewees at
the pharmaceutical and both paper mill companies considered it least important. One
interviewee at the medium petrochemical company argued that energy efficiency is only
considered during the call for tender, since it is based on theoretical values and calculations.
The interviewee then continued, saying that the company is committed to use energy efficient
engines. It was stated by the micro beverage that energy consumption is not one of their
largest costs and is thus not prioritized. The interviewee at the pharmaceutical company stated
that compressors are, by definition, consuming large amounts of energy. Energy efficiency is
thereby related to the operations of the machines.
5.2.6 Purchase factors – Future classification
The future desired characteristics of compressor offerings were structured based on the same
six factors as presented in figure 5.4. The interviewees stated how they perceive the relative
importance within 5-10 years with a motivation to their opinions. Discussions of how the
customers are likely to develop their business in general followed since it was considered by
the interviewees that such development might have an impact on the use of air compressors.
36 |
Chalmers University of Technology | Figure 5.4. Illustration of the relative importance of the purchase factors within the next 5-10 years
During the discussion of how the customers’ needs would change within the coming 5-10
years, five out of nine interviewees discussed energy consumption (see figure 5.4). Neither
interviewees at the medium petrochemical nor the small petrochemical company (1) believe
that energy efficiency will become of higher importance compared to the other factors in the
future. One interviewee at the small petrochemical company (1) argued that they would move
on to renewable energy sources and start reuse and recycle their waste to become more energy
efficient. Both the medium and small sized petrochemical companies had the same thoughts
but discussed further the possibilities of altering the production process into using renewable
raw materials such as bio oil instead of naphtha. Both also stated that compressed air is
something that they would avoid using from an energy perspective. Compressed air requires a
lot of energy, which is getting more expensive, according to one interviewee at the small
petrochemical company (1). Interviewees at the large paper mill company would also prefer
not using compressed air, if it could be avoided.
Discussing energy efficiency, representatives for the large paper mill company and the micro
beverage believed it to be of greater importance in relation to the other factors. The reason for
that choice is that the production process will be developed more in the future and therefore
the energy consumption will have an increased impact on the business. The micro beverage
would be likely to invest in water cooled compressors in the future, to recycle the heat and re-
use it in their production. Interviewees at the large paper mill company on the other hand
made a connection between energy efficiency and quality of the products. The interviewee
further expressed an interest in updated control units for more efficient use of the machines.
Interviewees at the heavy automotive company (1), the pharmaceutical company and the
small petrochemical company (1) also desire upgraded control units on the compressors to
increase energy efficiency and matching the air production with the need. Both the large paper
mill company and the large pharmaceutical company specifically desire upgraded frequency
controlled machines but also superior control units. According to the interviewees, the
superior control unit facilitates interoperability by regulating all compressors based on air
consumption. The heavy automotive company (1) discussed installing such control units in
the near future and are in the process of hiring one from ACCT.
37 |
Chalmers University of Technology | When asked to describe how the perception on service is likely to change, three out of nine
respondents considered that service offerings would be increasingly important onwards. The
ability for the supplier to wirelessly be connected to the machines, to monitor their
performance and to store information was also discussed. Both paper mill companies had a
positive attitude to ACCT in both cases, monitoring and communicate with the machines.
According to the medium paper mill company an intelligent communication system would
facilitate maintenance and decrease lead time, since the physical inspection in many cases
may default. One interviewee at the medium petrochemical company on the other hand was
skeptical to the concept of ACCT communicating with the machines in real time. All
maintenance has to be scheduled since the machines operate around the clock providing air
for the production. Therefore would a communication system be useless according to the
interviewee. The interviewee continued saying that a system for surveillance and diagnosis of
the machines would be of greater interest since it would give an indication of possible failures
in advance.
In addition to the discussions regarding the factors, one interviewee at the medium paper mill
company speculated that compressor suppliers might start using a similar system as Siemens.
The interviewee stated that customers charge for placing orders, instead of the customer doing
it themselves using a catalogue. The interviewee continued saying that to simplify the
purchase of spare parts, suppliers could synchronize the product names by using different
standards valid for different industries. Respondents at both the medium and the small
petrochemical company (2) are having a negative viewpoint on decreased personal contact
due to digitalization. The argument was that both during the purchase process and when
contacting service center for maintenance, human interaction is desired.
Customers’ business development
The micro beverage, both small petrochemical companies, the medium petrochemical
company and the heavy automotive companies will increase capacity in their production and
at the same time increase efficiency by doing more with less effort. Possibilities for increasing
automation to decrease manual labour in the production process has been recognized at the
micro beverage, while both heavy automotive companies and the small paper mill company
are adapting into using digital tools. The small paper mill company is developing towards
incorporating IoT into their production system, allowing producing units to communicate
directly with each other. The benefits of such advanced and highly automated system is the
decreased dependence on individual’s competence, according to one interviewee at the small
paper mill company. When the production process are becoming increasingly sophisticated at
the heavy automotive companies the need for properly trained staff increases, resulting in
longer training periods. To reduce the risk of “the human factor” a pick to light system is used
to ensure the kitting operations are performed correctly. The next step in digitizing the
production for the company is starting to use digital supporting documents, related to each
product.
Five out of nine respondents expressed that when increasing capacity, the air consumption
will increase. The cost per cubic meter produced air is of great interest at the medium
petrochemical company. It was mentioned at interviews with both the medium petrochemical
company and heavy automotive company (1) that there are difficulties in measuring energy
efficiency and air consumption. One interviewee at heavy automotive company (1) argued
that installing measurement equipment would itself affect the consumption negatively, since
38 |
Chalmers University of Technology | the equipment would disturb the airflow. One company that have chosen to start measuring air
consumption is the large paper mill company.
5.2.7 Competitors
The global air compression industry is characterized by a few large players with strong brands
that frequently acquire smaller companies and engage in strategic alliances (Technavio,
2015). The rate of know-how is typically high and companies competing globally spend a lot
of resources on development and innovation (Technavio, 2015). Three companies was
selected for the study, Ingersoll Rand represents a large incumbent while Kaeser and ELGI
represent aspiring smaller companies (Technavio, 2015).
Kaeser has recognized the customer need of compressed air rather than the compressor itself,
now developing network solutions for customers to buy compressed air by use (Schulete,
2016). In addition, Kaeser is also developing predictive maintenance systems for their
compressors and are planning to create paperless factories where all information is digital
(ibid). Ingersoll Rand on the other hand has chosen a different approach, but still based on
customer needs (Szefur, 2016). The company uses its expertise in different industries to
identify the part their compressors is taking in a larger system, to ensure that the compressor
is contributing to an optimized system overall and thereby create customer value (ibid). ELGI
has recognized the industrial trend of IoT and Industrie 4.0 but are currently mainly focusing
on expanding its business globally (Nanda, 2017).
In the following table, the competitors are presented by their value proposal, digital
development focus and competitive strategy, see table 5.2.
Company Value proposal Digital development focus Competitive strategy
Ingersoll Embrace sustainability and Energy efficient systems Energy efficient solutions
Rand2,3 emphasize energy Deliver data driven end-to-end
efficiency More customized solutions systems assessments
with use of data driven
insights and analytics
Kaeser4,5 Emphasize both their own Predictive maintenance Low life cycle cost
and customers CO2 system Energy efficient solutions by
consumption superior compressor technology
Offering various financial
alternatives to increase availability
ELGI6 Offer products to give Predictive analytics Forecasts to become the best at
continuous uptime energy efficiency to give value for
Energy efficient services the money to the customer
Ensuring energy efficiency
at customers
Table 5.2. A summary of the focal company’s competition.
2 Company Ingersoll Rand (2017)
3 Szefur (2017)
4 Kaeser kompressoren SE (2017)
5 Kaeser (2017)
6 Nanda (2017)
39 |
Chalmers University of Technology | 6. Analysis
In the following section, the nine different building blocks of a Business Model Canvas will be
analyzed with the studied company, ACCT, as starting point. From the Business Model
Canvas analysis, three proposed business models will be presented, explained and motivated.
The proposed business models are thereafter discussed with regards to efficiency and
profitability in mind. The section will be concluded with an action plan of how to transform
successfully.
6.1 Business Model Canvas analysis
The Business Model Canvas analysis aims to investigate each of the nine building blocks that
Osterwalder and Pigneur (2010) present in their Business Model Canvas. In line with their
division of the nine building blocks mentioned in section 3, Frame of reference, the analysis
below will be divided into the following parts; customer, offer, infrastructure and finances.
From the internal empirical results, it has become clear that ACCT is well aware of the
technological changes to which they are exposed to. As written in the same section, ACCT is
continually working on adopting new technologies that can be advantageous for their
competitiveness. The difficulty lies rather in the adoption of the technological change which
goes in line with Wessel et al. (2016) discussion of the 20th century powerhouses. Both
Westerman et al. (2014) and Heaslip (2015) discuss the two dimensions of a successful
transformation; the digital technologies and the culture and leadership skills. In the case of
ACCT, the bridge between digital technologies and a successful transformation seems to lack.
The empirical results highlights the fact that ACCT struggle with a lack of digital knowledge
and cultural aspects, such as willingness to transform.
6.1.1 Customers
As mentioned in the section 3, Frame of reference, the customer category involves the
following three building blocks; customer, channels and customer relationships. These
building blocks will be analyzed separately.
Customer
In order to define customer segments and understand the differences between them, one need
to understand their customers’ behavior and characteristics (Osterwalder & Pigneur, 2010). In
the analysis below, the following tools will be used to investigate ACCT’s customers; the
perception of technological adoption and the level of digital potential each sector possess. The
aim is to investigate which groupings of customer segments that are compatible with a
potential digital transformed value proposition. In the analysis below, conclusions about
industries will be drawn solely based the interviews that have been performed with the
represented companies.
Based on the customer's production needs, their latest developments and their digital
standpoint, see section 5, Empirical results, the respondent companies have been categorized
with reference to Rogers’ theory of how new technology are adopted.
40 |
Chalmers University of Technology | Figure 6.1. The customers in Roger’s diffusion of innovation curve.
As seen in figure 6.1, there are no innovators categorized among the respondents. Based on
the empirical results, all the respondents operates within mature industries with little space
and incentives to innovate. The leftmost group are the early adopters where the heavy
automotive companies can be found. These respondents can be compared to opinion leaders
because of their self-powered innovative way of working, both physically in terms of
production process and mentally in their way of working. The two paper mill companies are
categorized as early majority together with the pharmaceutical and the micro beverage
company. Given the information in section 5, Empirical results, the companies are perceived
as innovative in the sense that they seem open-minded to adopt a new product after they seen
it useful by the early adopters. For example, the micro beverage company has seen how
beneficial it is to have automated production processes and therefore wish to have a more
automated production themselves.
All three petrochemical companies are placed on the right side of the curve. All three actors
represented more or less a sceptic mindset towards change and the prevailing digital
transformation. In some contexts, the medium sized petrochemical company also
demonstrated a resistance towards new innovative solutions and that they preferred living in
the past with traditional communication, which can be seen in the discussion where they
showed a skeptical mindset towards machine communication and automatization of various in
house activities. This particular respondent could be compared with Roger’s definition of
laggards.
Table 6.1. The studied industries’ digital potential.
In addition to the diffusion of innovation theory, a research made by Gandhi et al. (2016),
presented in section 2, Industrial digitalization, shows what process the industries have digital
potential within, see table 6.1. This classification is used as a tool when determining
customers’ potential to digitally transform. Among the industries mentioned by Gandhi et. al
(2016), three of them are aligned with the respondents; chemicals, pharmaceuticals, and basic
41 |
Chalmers University of Technology | goods manufacturing. All three sectors, pharmaceuticals, chemicals and basic good
manufacturing are classified by Gandhi et al. (2016) as capital intensive sectors with the
potential to further digitalize their physical assets. In addition, these sectors can also be
designated as business to business sectors with the potential to digitally engage and interact
with their customers (Gandhi et al., 2016). For basic goods manufacturing solely, Gandhi et
al. (2016) also characterize it as a labor intensive sector with potential to provide digital tools
to its workforce.
As seen from the above analysis, the result of the customer segmentation by Rogers (2003)
theory and Gandhi’s (2016) classification of digital potential among different sectors
correlates. Therefore, Roger’s (2003) and Gandhi’s (2016) theories will be used as a
combination in the analysis below. Table 6.2 presents the segmentation that is based on the
correlated theories mentioned.
Companies Roger’s theory Digital potential
Heavy automotive L Early adopter Physical assets
Customer interaction
Workforce
Heavy automotive L Early adopter Physical assets
Customer interaction
Workforce
Paper mill L Early majority Physical assets
Customer interaction
Workforce
Paper mill M Early majority Physical assets
Customer interaction
Workforce
Beverages Mi Early majority Physical assets
Customer interaction
Workforce
Pharmaceutical Early majority Physical assets
Customer interaction
Petrochemical S Late majority Physical assets
Customer interaction
Petrochemical S Late majority Physical assets
Customer interaction
Petrochemical M Laggards Physical assets
Customer interaction
Table 6.2. Summary of the customer segmentation.
42 |
Chalmers University of Technology | As seen in table 6.2, there are various levels of adoption rates as well as digital potential. As
for the companies that operates within the petrochemical sector, Gandhi et al. (2016) means
that there is potential for digital transformation in terms of physical assets and customer
interactions. However, they seem to lack the potential and will of providing digital tools to the
company’s workforce, which goes in line with the observations that have been found during
the empirical studies. According to Gandhi et al. (2016), the basic goods manufacturing
sector, namely heavy automotive, beverage and paper mill companies, have potential within
workforce, customer interaction and physical assets. Aligned with observations from the
external empirical results, found in section 5, these customers are receptive to transformation.
In line with Hammer et al. (2016), to reach a successful transformation it is important to
gradually change with an agile strategy. By targeting the actors with high digital potential,
Hammer et al. (2016) mean that the more hesitant parties can be reached after it has generated
understanding and excitement.
Customer relationship
Osterwalder and Pigneur (2010) distinguish between three customer relationships but in the
analysis below, only the relationships regarding current customers will be considered. As the
external empirical results indicates, the current customers value personal contact, while the
development goes towards a more automated ones (Wessel et al., 2016). Judging from the
empirical discussions, the customers think that an improved digital contact will damage the
close relationship to the focal company. Two of the petrochemical companies oppose a more
digital customer contact, both in terms of purchase process and service requests. Other
respondents are more attentive to a potential change, see empirical results.
The use of big data analysis can facilitate a more automated relationship with the customer
(Techtarget, 2017). By gathering information about the customer, ACCT can customize its
relationship and make them more personal. The information gathered can also be used to
proactively offer customers new personalized value propositions, a statement that is
recognized by the customers in the empirical results. In addition to a more customized
response, a more automated customer relationship would facilitate for ACCT in their
operations. As the General Manager said during the interview, less personal interactions in
some activities would create more value added in other activities.
Channels
From the empirical results, it becomes clear that the current customers want to be reached in a
rather traditional way, such as telecommunication and personal meetings. Although, there are
some respondents that are more open towards digitalization and its advantages than others. In
line with Wessel et al. (2016) argumentation that internet is found everywhere today, in
people's phones and homes, ACCT should take advantage of the opportunity that can enable
them to reach customers around the clock. As stated in the empirical results, section 5, ACCT
is currently working on a website where they can reach out to their customers in a more
digital way and use online communication as a complement to the personal communication
activities. This provides a foundation for further digitalization. In some industries, such as
fashion industry, social media channels are considered to be useful when delivering value
propositions to the customers (Westerman, 2014). Although, with the empirical results in
mind, both ACCT and the customers agrees on the fact that the compressor industry deliver
products and services for industrial customers and not for private individuals that are the main
users of social media channels. However, it can be discussed if social media channels could
become a successful channel in the future.
43 |
Chalmers University of Technology | 6.1.2 Offers
Opportunities for value creation have been defined by the customer demands and industry
trends from section 2, Industrial digitalization. In the creation of value propositions, Dave
McClure’s elevator ride template of what, how and why. The proposed offers will be
presented below.
Offer the right spare part at the right time to decrease inventory costs
Customers have indicated an interest in lowering the inventory costs related to spare parts.
The reason why customers store large amount of spare parts is originated in skepticism
towards the delivery precision of those parts, because of the high need for dependability. By
using sensor technology, wireless communication and big data analysis, the life of spare parts
can be predicted which results in right spare part arriving at the right time, not at the expense
of dependability.
Offer right product at the right time to increase dependability
The customers have shown an increased interest in suppliers that are dedicated to the overall
use of compressed air. By predicting the life of the machines and mapping the customers
demand for compressed air, the focal company can approach the customer proactively with
offerings for additional compressors, in the case of an increasing demand for air and also offer
new updated machines to replace old ones.
Offer right service at the right time to increase the machine’s dependability
Customers press the importance of dependability of a machine due to the fact that their
production depends on compressors’ functioning. In line with these observations, other actors
have also recognized the need of proactive maintenance due to more complex and
sophisticated production plants. The machines condition and use are analyzed to predict the
life of the machine and its components in order to provide tailored service execution.
Offer more time-efficient services performed by technicians to decrease service costs
In the empirical results it has been shown that customers have opted out options because of
high service costs and is thereby an influencing factor when choosing supplier. More efficient
services can be performed by technicians geared with the information and knowledge needed
to perform the job. In order to be prepared, the service technicians could be more connected,
have better access to information and real time data and communicate online.
Offer energy efficient operations to decrease total cost of ownership of the machine
The empirical result indicates an increased importance of energy efficiency within five years
with the argument that the consumption is increasing and energy is getting more expensive.
By modular programming, flexibility is achieved which will ensure efficient operations of all
machines and decrease the cost of use and thus of long term ownership. In addition to a
decrease in total cost of ownership, the value proposition may also demonstrate environmental
responsibility externally.
Offer air as a service where customers pay by use which in turn decrease costs
In the empirical results, alternative ways of getting compressed air have been discussed.
Offering air as a service has been seen as a requested option as the customer then only pay by
use, cost per cubic meter, and thereby escapes the fixed cost for a compressor and the
associated costs for maintenance. In line with the customer request, traditional product-based
companies are now moving towards servitization.
44 |
Chalmers University of Technology | Offer cheaper products by improving product and process development
The empirical results demonstrate the importance of price when choosing compressor
supplier. In those cases when the focal company has been deselected, it has been due to price.
With transparency, use of today’s technology, the development can be streamlined and
decrease the cost of producing a unit. By using visualization tools and digital communication,
the customer can be involved early in the process, regardless of time and place.
6.1.3 Infrastructures
In the following section the infrastructures that are required to provide the suggested
proposals are presented. The area of infrastructure includes the following building blocks;
key activities, key resources and key partners.
Key activities
The key activities presented in this section correlates with the presented value propositions in
the above section. The activities required for product and process development, efficiency and
availability of maintenance and service, energy efficiency, predicted maintenance and air as
service will be presented.
Product and process development
Judging from the discussion presented in section 5, Empirical results, many customers base
the purchase decision on the total cost of owning the machine including the initial cost of the
product. A consequence of a streamlined development process is, according to Ratkovic
(2013), decreased product cost which may be reflected in the offered initial price. In
streamlining the development process, adopting VR technology has been proven to be
particularly suitable according to Pease and Zistl (2014). By using VR tools during the early
stages of development for simulating testing instead of performing physical ones, costs of
labor and hardware can be decreased, as described in section 2, Industrial digitalization, The
ability to perform tests early in the process increases the probability of detecting problems
early in the process which will make them less costly (Nigel & Slack, 2015). Another way of
ensuring that problems are handled early in the process is engaging customers in the concept
development phase of the process according to Baur and Wee (2015).
Efficiency and availability of maintenance and service
Service has proven to be one of the most important purchase factors for customers (see
section 5, Empirical result) and the lead time on services is in many cases a key for assuring
dependability (Ratkovic, 2013). It has been noted in the Empirical results that the physical
availability of a service technician lowers the cost of the service, which is a desired feature for
customers as well. Even though the service technician is physically available Powell (2016)
has recognized a lack in efficiency due to missing knowledge, requiring technicians to travel
back and forth from the site. In order for ACCT to avoid such issues, providing service
technicians with information and communication devices such as smart gadgets will
streamline the service activities (Powell, 2016). With proper use of the gadget, the technician
can perform a high quality service efficiently on site and thus increase dependability on the
machine while decreasing the costs of the service (Powell, 2016)
Predictive maintenance, spare parts
According to customers, spare parts are bought and stored in order to ensure dependability of
the machine (see section 5, Empirical results). By digitally analyzing the life of the machine
and its components, precise indications will be provided on when a certain component needs
to be exchanged in advance and when a certain service needs to be provided (Siemens, 2013).
45 |
Chalmers University of Technology | The information can further be used by service technicians both to schedule services when
needed and order spare parts which arrives right on time (Siemens, 2013). In providing such
offering which is referred to as predictive maintenance, the dependability of the machine will
increase and at the same time, customers and ACCT´s cost of spare part storage will decrease
(Siemens, 2013). Predictive maintenance is made possible by the combined use of sensor
technology, wireless communication and big data analysis to transmit, gather and analyze
information in an IoT system (Bossen & Ingemansson, 2016).
Predictive maintenance, proactive sales
The ability of suppliers to support the use of compressed air is being requested by customers
(see section 5, Empirical results). Such ability will be provided by the combined use of sensor
technology, wireless communication and big data analysis. The system will enable analysis of
the life of the machine and the demand for compressed air over time (Techtarget, 2017). In
understanding the condition of the machines and changes in compressed air demand,
customers can proactively be approached with customized offerings for compressor
exchanges, upgrades and new products (Techtarget, 2017). In the case of ACCT, the sales
department needs to be able to understand the data received.
Energy efficiency
Energy efficiency was considered to become an important purchase factor in the future (see
section 5, Empirical results) and because of the policy changes towards sustainability (see
section 2, Industrial digitalization) ACCT will provide energy efficient solutions. Since
customers also expressed a need for personalized offerings ACCT will, by utilizing modular
programming in combination with technical superiority and efficient operations, meet both
demands. According to Golfman and Lammers (2015) modular programming implies easy
updates of the system controllers’ software on request. Modularized software enables ACCT
to offer customers standard products, differentiated by the software modules. The modules
will differ from each other in the level of complexity and thus the ability to provide energy
efficient operations. The modules can easily be changed since the machines are wirelessly
accessed by ACCT. In order to successfully implement modular programs, the right
competence is needed.
Air as service
Customers has expressed a possible future interest in pay by use solutions for compressed air
and personalized solutions (see section 5, Empirical results). In addition, Baur and Wee
(2015) have recognized business models characterized by pay by use offers to be favorable. In
order for ACCT to estimate the amount of air the customer have used, the demand for air has
to be measured by ACCT themselves. This activity requires both knowledge, time and
resources in order to reach a desired result. In addition, as the value proposal change in terms
of ownership of the machine, service agreements with customers will be diminished. Instead,
a pay by use agreement will be established with the customers and ACCT will be responsible
for the machine and its maintenance.
46 |
Chalmers University of Technology | Key resources
According to Osterwalder and Pigneur (2010), key activities can either be of physical,
intellectual, financial or human character. In the below analysis, only the physical and human
resources will be discussed due to the chosen scope of the thesis.
Physical resources
To meet the increasing demand for energy efficiency expressed by the customers, ACCT offer
system controllers for efficient operations of their machines by interoperability (see section 5,
Empirical result). To ensure that the systems are robust and reliable, in accordance with Wee
et al. (2015), ACCT can use standardized systems for all sub control units. By offering robust
and reliable operations systems, ACCT also creates customer value in the form of
dependability. Dependability was described by the customers in section 5, Empirical results,
as superior to quality, service and lead time of service. The one customer using leasing
contracts expressed difficulties in integrating control systems, especially in the case where the
machines had different owners (section 5, Empirical results). Efficient control of the
machines is related to energy efficiency according to the customers and will be increasingly
influential in the future (see section 5, Empirical results). As described in section 2, Industrial
digitalization, flexible systems may be achieved by transparency between all owners of the
software and machines (Wee et al., 2015). As transparency is a characteristic desired by the
customers (see section 5, Empirical results) ACCT do not only provide customers with
increased energy efficiency, such customer value also reinforces the relation.
Using common standards on control units facilitates larger wirelessly connected systems of
systems which constitutes for internet of things, as described in section 2, Industrial
digitalization. Having machines wirelessly connected to ACCT, continuously transmitting
information of use and condition requires resources and tools to handle the incoming data
(Davenport & Dyché, 2013). To successfully manage the data and utilize the information,
resources for data management to ensure high quality data, data mining to recognize patterns
in the flow of information, Hadoop to store data, in memory analytics for immediate analysis,
predictive analytics to statistically predict likelihoods of outcomes and text mining to by
machine learning analyze written texts are required. The combination of these resources and
robust systems, the focal company can predict the life of components and demand and thus
proactively offer solutions and dependable machines (Ratkovic, 2013). Such offerings are
expected to be appreciated since customers have expressed a particular interest in suppliers
proactively offers additional value based on knowledge (see section 5, Empirical
results).
Another prerequisite for successful use of big data analysis regards the components
transmitting the information wirelessly as discussed in section 2, Industrial
digitalization. Since customers demand energy efficiency and low lifecycle cost, see section
5, Empirical results, these components must align with this need as well. In order to maintain
a low price on the product during its life, the components must be low priced but still be of
high quality, have a long battery life and low energy consumption (Bossen & Ingemansson,
2016).
The use of such gadgets implies large amounts of background activities. The smart gadgets
themselves are only meant to provide the user with relevant information for the task and
abilities for limited interaction (Powell, 2016). In offering predictive maintenance services,
technicians would be provided with real time information wirelessly transmitted by the
machines sensors through a internet of things (Ratkovic, 2013). Information are shown
47 |
Chalmers University of Technology | visually to the user by the gadget and for interaction, the gadget may contain a recording
device with voice recognition, a camera or scanner (Ratkovic, 2013). The gadget will
contribute to the value creation for customers by simplify and streamline the tasks performed
by the user (Ratkovic, 2013).
Human resources
In addition to the physical resources ACCT needs to possess in order to deliver a value
proposition to the customers, human resources are as important in a successful transformation
as physical resources (Osterwalder & Pigneur 2010). Human resources have a tendency to
even be forgotten sometimes (Cleverism, 2015). In the empirical results, section 5, personnel
and human interaction are emphasized in every stage of the customer relationship, both by
ACCT and the respondent customers. Regardless of value proposition, human resources are of
great importance to ACCT. As stated in the empirical results, the customer is not only buying
a compressor, they are buying the ACCT’s experience.
In order to reshape a value proposition to one with digital character, the employees need to
conform. Since previous digital projects, such as the implementation of SAP see section 5,
Empirical results, resulted in a longer learning curve than expected, the human resources are
of even greater importance to ACCT this time.
As discussed by Westerman et al. (2014) in section 3, Frame of reference, the barriers to
transform often involves lack of leadership, governance, culture and financial cases. This
reasoning corresponds with the empirical case where ACCT describes an earlier digital
implementation with the same transformational barriers. As written in section 3, Westerman
et al. (2014) discuss the following four element how to succeed with a digital transformation:
possess a common vision, an engaged organization, an improved governance and good
leadership. In the case of ACCT and in line with Westerman et al. (2014) suggestions, the
following resources and activities, regardless of value proposition, are recommended in table
6.3.
Resource Activity
Vision Software Familiarize ACCT with the proposed technologies by
engineers using workshops
Engagement Project managers Create solutions together with employees by using
crowdsourcing
Governance CEO Put someone in charge for transformation
Leadership Knowledge pool Build up the right digital skill
Table 6.3. A presentation of the suggested human resources and corresponding activities.
48 |
Chalmers University of Technology | Key partners – Partner networks
In order to deliver a value proposition, Osterwalder and Pigneur (2010) believe in partner
networks. As discussed in section 3, Frame of reference, there are different types of
relationships which can be applied at the ACCT case. In the empirical results, the employees
at ACCT explain that they are skeptical towards partnerships. However, from the external
point of view, the customer side, it is preferably desired. The customers emphasize a total
demand that can be met by one supplier, as discussed in section 5, Empirical results. To
achieve this, partnerships are required. As understood, a collision occurs between ACCT’s
mindset and the customers’ demand.
As mentioned above, Osterwalder and Pigneur (2010) discuss four different types of
partnerships, one of them is strategic alliances where companies can benefit from others’ core
competences. Wessel et.al (2016) discuss the Industrialist dilemma and its difficulties to
change well established internal systems and they mean that the powerhouses of 20th century
that still operates today have a difficult time adapting to this new data driven world. In line
with the above reasoning, the Industrialists could benefit from strategic alliances with
companies that possess digitalization as their core competencies. It can give better access of
knowledge and increase effectiveness by producing better products and services (Wessel et al.
2016). In the case of ACCT, partnerships with software developers, customers and technology
vendors could enhance their competitiveness due to additional knowledge.
The remaining two partnerships; Coopetition (cooperation between competitors) and joint
venture, actors that creates new businesses will not be discussed due to ACCT’s core values.
6.1.4 Finances
The section consists of the two building blocks: cost structure and revenue stream. In the
analysis below, these two will be further discussed with regards to the suggested value
propositions.
Cost structure
When adopting digital technology initial costs related to implementation and adoption appears
(Bossen & Ingemansson, 2016). Costs of investing in new equipment and installation appear
at the moment of purchase but there are also other following costs (ibid). Implementing new
technologies require education in the use of the new technology, in order for the users to make
the initial investment worth it. (ibid). Automation of activities that would otherwise been
performed by individuals, labor costs decreases and remaining personnel can focus on value
adding activities (Nigel & Slack, 2015). In the case of ACCT, this create a reduction in labor
costs.
Implementing digital technology in the product development and production planning process
will include costs of implementation, adaptation and adoption (Bossen & Ingemansson,
2016). The cost of producing products will decrease when the new technology is installed
(ibid). The product cost will decrease since the new technology will imply shortened
development time and thus less development costs, which are imbedded in the total cost of
producing the product (Nigel & Slack, 2015). By adopting digital technology during the
production planning stage the trial time for producing the new product and including it in the
production logistics decreases which implies a smooth implementation and thus decreased
costs (Bossen & Ingemansson, 2016).
49 |
Chalmers University of Technology | Since a majority of the people employed is related to service, service operations have a large
effect on the overall cost of ACCT´s business (see section 5, Empirical results). It has been
recognized by Powell (2016) that service technicians often lack time efficiency. More time-
efficient service technicians imply more tasks performed on the same amount of time and thus
the cost of labor for each task will be lowered Powell (2016).
Revenue stream
As specified in section 1, Introduction, ACCT aims to increase the company’s profitability
and one way to do that is to increase the company’s turnover/revenue stream. Osterwalder and
Pigneur (2010) distinguish between one-time customers and continuous payments from a
customer segment. In the following analysis, continuous payment will only be discussed due
to the formation of customer segments that is solely based on the current customer base. The
chosen value propositions will create a variety of revenue streams. Initially, a differentiated
offer will create a totally new revenue stream for ACCT and thereby incur revenue. In
addition, recurring revenues occurs in conjunction with those value proposals that are likely to
continue in the future, such as the service based value proposals. The service based proposals
also include value added services in terms of more activities performed in less time. ACCT
will also be able to charge more for those value added services. In line with the discussion
above of a potential website development, advertising in a greater extent could be a possible
revenue stream for ACCT. An extended website will attract companies that want to pay to be
visible at their website and therefore create a revenue stream from advertisement.
6.2 Proposed business models
From the above analysis, three potential business models have been developed with the
grounds that those offers are assessed to be relevant for ACCT. The purpose of these business
models is to create added value for the customers and increase ACCTs competitiveness on the
market. The offers that were rejected are not considered to be sufficiently demanded by the
customers nor a competitive mean in the compressor industry, based on the empirical results
and the findings on industrial digitalization. All offers can be found in section 6.1.2 Offers
and the ones that was chosen to be part of the business models was “Offer energy efficient
operations to decrease total cost of ownership of the machine” which is found in Proposal 1,
“Offer right service at the right time to increase the machine’s dependability” and “Offer
more time-efficient services performed by technicians to decrease service costs” which are
both found in Proposal 2 and finally “Offer right product at the right time to increase
dependability” which is found in Proposal 3. These offers were each chosen since they were
the ones that are directed towards the main customer demands that were found during the
customer study (see section 5, Empirical results): dependability and energy efficiency.
Dependability was considered to be one of the two customer needs to focus resources on since
it is related to the three purchase factors that was considered to be of highest importance
today: quality, service and price. Energy efficiency was chosen since it was the one factor that
was considered to become of higher importance for the interviewed customers in the future.
The “Offer the right spare part at the right time to decrease inventory costs” offer was rejected
since it was not considered to affect the dependability of the products compared to the offers
that was directed towards service. The reason why the offers named “Offer air as a service
where customers pay by use which in turn decrease costs” and “Offer cheaper products by
improving product and process development“ were rejected were the lack of interest amongst
customers in both offers. Customers was willing to accept a high price for products of high
quality and the interest in buying air by use was currently not considered to be high enough.
50 |
Chalmers University of Technology | The customers have been segmented based on Roger’s diffusion of innovation theory and
Gandhi et al. (2016) discussion of industries’ digital conditions. As presented above, the
analysis touch upon four out of five consumer groups which correlates with the categorization
of digital potential and thereby, the segments will only be refereed as the four adoption groups
as continued, see table 6.4. As discussed in section 3, Frame of reference, it is useful to target
one group of customers and then gradually progress to others. In this case, in line with
Roger’s (2003) reasoning, it is recommended to begin with the early adopters due to their
technological adoption rate and digital potential, and gradually continue to early majority
when the product or service has been proven to be successful. As a result of this, it is
recommended to keep the old business models until the new ones are well-functioning for all
customer segments.
Early adopter Early majority Late majority Laggards
Heavy automotive L Paper mill L Petrochemical S Petrochemical M
Heavy automotive L Paper mill L Petrochemical S
Beverages Mi
Pharmaceuticals L
Table 6.4 Customers segmentation based on Rogers’ (2003) and Gandhi’s (2016) theories.
Proposal 1: Energy efficiency
From the given proposals, it has been understood that energy efficiency is, in addition to a
customer demand, a potential competitive mean. As seen in section 5, Empirical results,
energy efficiency has increased in importance for compressor companies around the globe. By
offering an energy efficient solution, the customers could reduce cost for energy consumption
at the same time as ACCT could gain value added sales, due to the new value offered to the
customers of measuring energy efficiency. From the empirical results, section 5, it has been
recognized that ACCT currently offers a parsimonious service for measuring energy
efficiency. This will probably facilitate the acceptance of the new offer. In addition, this
proposal can be seen from an environmental point of view. As mentioned in section 2,
Industrial digitalization, the demand for environmentally friendly products have increased in
conjunction with Agenda 2030 where development and environment are considered jointly.
The proposal can thus be seen as both advantageous from a cost and environmental point of
view, in line with Porter and van der Linde (1995).
The energy efficiency proposal involves an agile solution where the customer is able to scale
up and scale down the consumption digitally. Through modularization, together with system
controllers and wireless communication, the customer gives the possibility to add and remove
software functions. This simplifies programming updates in that sense that changes can be
done with as little effort as possible as discussed in section 2, Industrial digitalization. In the
prevailing industrial climate, actors need to be agile and modular to thrive in today's rapidly
changing environment (Fettke et al., 2014). With regards to that reasoning, a modular
proposal can increase ACCT’s competitiveness. The developed Business Model Canvas is
presented below, see figure 6.2.
51 |
Chalmers University of Technology | Figure 6.2. Illustrates the business model of “Energy efficiency”
To deliver a differentiated value proposition, close and personal communication between
supplier and customer is desired by the customer, see Empirical results section 5. Direct sales
is thereby a useful channel to communicate the provided value to the customer segment. In
addition, 24 hours service is a necessity to support customers with the new delivery. To
deliver a new value proposition which entails a new revenue stream, efficient sales
management becomes a key activity. The sales engineers need to know what they are selling
and to whom.
As explained above, new technological sets up also require human resources. To be able to
perform good sales activities, a well-established sales management is essential. There is also a
need for IT/software knowledge that can execute the software installment needed. To be able
to use the system controller, standardization of all units are also needed (Wee et al., 2015). To
make this business model effective, it is also recommended to have partnership, see figure
6.2, in order to support the more continuous relationships. Also, the customers are perceived
as a partner in this particular case as they will become co-developers.
Proposal 2: Offer efficient services at the right time
After a product has been sold, the opportunity to maintain it arises (See section 5, Empirical
results). It has also been proved that customers consider service as one of the highest valued
features when buying a compressor, which have been demonstrated by rejection of inefficient
services. In order to capture those business opportunities it is concluded from the customer
interviews that the right service needs to be offered at the right time. The opportunities does
not appear exclusively for ACCT but for competitors as well and customers choose the most
beneficial offer (see section 5, Empirical results). To provide customers with a superior offer
a proactive and efficient approach is proposed. In line with this reasoning the two value
propositions shown in figure 6.3 were combined to form the basis for this business model. By
combining the two value propositions (offer more time efficient services performed by
technicians and offer the right service at the right) both dependability and cost reductions can
be offered to the customer. In order for technicians to gain access to the right information, the
52 |
Chalmers University of Technology | use of smart gadgets are proposed. The purpose of the gadget would be the interaction, to
serve as a user interface (Powell, 2016). The information accessed by the gadget would
originate from the predictive maintenance system, to further utilize it for more efficient
services. By providing more efficient services, revenues from services will increase since
technicians will have the ability of generating more value for the customers with the same
amount of effort (Powell, 2016). This business model also implies increasing automation in
the back office, minimizing the human interaction which leads to lower labor costs and the
dependence on the single person's competence at the focal company (Pease and Zistl, 2014).
In offering proper services on products increases the life of the product, ensures a stable and
prolonged revenue stream from service.
Figure 6.3. Illustrates the business model of “Offer efficient service at the right time”
As figure 6.3 presents, the business model is characterized by big data analysis and the use of
smart gadgets. As big data analysis enables an extended interface between the human and
machine, see section 2, Industrial digitalization, the customer relationship will be almost
completely automated. The channel would be direct sales in order to still maintain a close, yet
still automatic, customer relation (see figure 6.3). Since the service opportunity will be
automatically predicted, the customer can gain access to that information automatically and
get the opportunity to order the service in advance via an online platform. The digital platform
on the company's website would contain a customer and a service technician interface. Both
parts would be able to log into the website and thereby gain access to schedules and relevant
information regarding the service operation. By having a trafficked website, the possibility for
revenue from advertisers marketing themselves on the website appears. To increase revenue
from advertising, the concept of improved pricing can be used match advertisers with
customers (LaRiviere et. al (2016). By accessing the real time stream of digital customer
information using wireless connections and big data analytics, ACCT can gain objective
information about the customers (ibid). In order for the service technicians to handle gadgets
needed to perform the service efficiently they need to be trained. The education’s aim is to
make the service technicians adopt the technology.
53 |
Chalmers University of Technology | Proposal 3: Offer right product at the right time
Given the rapid transformation of the industry sector, ACCT needs to proactively capture
business opportunities to increase revenue. With access to a flow of real time data of the
machine, ACCT can approach customers with a new offer when their current machines is
close to collapse. Proactive sales management can be achieved with the use of predictive
analytics to predict the demand (Davenport & Dyché, 2013). In addition, the use of wireless
communication reduces the cost of devaluing data due to less human work and decreased
labor costs.
In addition, wireless communication is often used to reduce cost to devalue data. As described
in Proposal 1, in order to succeed with proactive sales offers, good leadership is important at
the focal company. By measuring the condition of the machines, the sales engineers can
recognize a need even before the need has arisen at the customer. Since the customer is
approached before the need is realized, competitive offerings can be blocked in advance,
preventing competitors to increase their business. The sales engineer can be at the right place
at the right time. This value proposition will create more opportunities for ACCT which in
turn can be transformed to sales and further increased revenue. The customer will gain a
continuous flow of products and reduce the risk of potential breakdowns, namely an increased
level of dependability.
Figure 6.4. Illustrates the business model of “Offer the right product at the right time”
As seen in figure 6.4, the value proposition will be communicated in a semi-automatic way
which means that it will be automated where it is possible. Direct sales will be used as a way
to approach customers with the supplement of an extended version of today’s website. A
combination of personal and digital channels is recommended due to the predicted
maintenance system along with the customer demand. As mentioned above, in order to reach
a predictive maintenance system, tools for predictive analytics and a wireless communication
system will be needed. The required technologies needs to be installed and lectured in how to
use in order to make the transformation as efficient as possible. In addition, the sale engineers
need to learn how to interpret the data correctly, creating another need of education. A
54 |
Chalmers University of Technology | strategic partnership with an IT consultancy company would be necessary in order to support
the new installments and required training.
6.2.1 Discussion of the proposed business models
In order to reach a sustainable transformation, improvements of both existing products and
services are desired (Westerman, 2014). As mentioned in section 3, Frame of reference,
Bilefield (2016) emphasizes the importance to transforming throughout the organization in
order to succeed with a digital transformation. In line with these reasonings, the three
proposals presented above will be linked together to create a completeness and sustainable
transformation. The three proposals together represent an improvement in both product and
service and as they involve both sales, product and service department, the transformation will
be performed throughout the organization. The digital transformation of ACCT can be
illustrated with a circular orbit, see figure 6.5.
Figure 6.5. The proposed process of digital transformation.
As the discussion above proposes, the linkage between the three proposals can be presented as
following. Initially, ACCT should offer an energy efficient solution due to future customer
demands and external forces such as increased cost of energy and environmental aspects. In
order to maintain the condition of the machine, efficiently performed services are required.
With a proactive approach, the customer will receive time efficient service at the right time
without any disruptions in production. Someday, the machine will need to be replaced. Using
proactive sales management, ACCT can offer the right product at the right time, prior the
customers’ awareness. In line with the first proposed business model, the right product is
characterized by energy efficiency. In order to reach a sustainable transformation, all three
proposals in the circular orbit is of importance. As seen in previous section, all three business
models have some resources and activities that are alike and those will be shared between the
models.
As discussed in the above analysis, the customers are segmented into their digital potential
and adoption rate. In line with Roger’s (2003), the heavy automotive companies were
classified as early adopters and the medium sized petrochemical company was classified as a
55 |
Chalmers University of Technology | laggard. Since the characteristics differ between the early adopters and the laggard, the total
duration of the circular orbit will be shorter for the early adopter compared with the laggard.
As mentioned in the external empirical results, the medium sized petrochemical company
does not consider that the importance of energy efficiency will increase in the future. Those
actors will thereby be more difficult to persuade why they should buy an energy efficient
solution. In addition, as laggards tend to be very traditional in relation to new proposal, the
process of demonstrating the benefits of a digital product or service will be longer in
comparison to the heavy automotive companies that are perceived as open and accessible for
new opportunities. It may thus be considered unprofitable to put effort and allocate resources
on digital offerings to the laggards.
Since these three business models all contain digital elements, combined with the fact that
ACCT does not yet possess all digital capabilities needed, the digital adoption process is of
great importance. Because of the high importance of digital adoption, it is also considered to
be the main obstacle for ACCT to successfully implement these business models and linking
them together.
6.3 Action plan for ACCT
In order for ACCT to exploit the benefits of digitalization and to utilize the suggested
business models, ACCT should, in line with Libert et al. (2016) and Maxwell (2017)
framework of transformation, ensure the following steps in the digital transformational
process:
Reality check
In order to reach a successful transformation, ACCT needs to clarify their starting point in
terms of what they do and how they do it. In the case of ACCT, they should be aware of how
common changes are executed within the company and use that as a starting point, see section
5, Empirical results. While documenting these activities, it is also important for ACCT to
realize that what they did years ago might not be the most efficient way to do it today. For an
incumbent company, realizing its inferior position in digital contexts is a first step. Since the
powerhouses of 20th century have difficulties in adapting to new digital initiatives, as Wessel
et al. (2016) argue, it is essential for ACCT to realize that they might ask for help and
direction from external partners. The internal empirical results demonstrate a strength in
technological awareness while a weakness in actual usage and mindset towards change. As
previous digital projects have indicated, the obstacles with implementation have been the lack
of adoption. In order to avoid previous mistakes, ACCT should as discussed above clarify
their starting point. ACCT appear confident in their own digital abilities, stating that they
have nothing to gain from collaborating with other companies when developing new
technologies (see section 5, Empirical results). Such over self-confidentiality implies the risk
of not being attentive enough to their surroundings, for ACCT, and thus the risk of being
overrun by innovative 21´th century corporations increases (see section 2, Industrial
digitalization). For this reason, it is of high importance for ACCT to truly engage in this first
step and to avoid being subjective when clarifying their starting point.
In addition, Maxwell (2017) stresses the importance of understanding the customers’ and their
total needs. As seen in the internal empirical results, customer differs in requirements,
production needs and what they value. In line with this, ACCT might consider changing the
assumption of product based segmentation and rather use a more customized approach, as the
segmentation presented above.
56 |
Chalmers University of Technology | Company assets
The second step in achieving a successful transformation, is according to Libert et al. (2016)
to do a compilation of all assets within the company, tangible as intangible. In line with
Osterwalder and Pigneur (2010) reasoning about the oversight of human resources, it is even
more important for ACCT to emphasize the intangible assets such as skills of employees. In
the empirical results, previous digital project are presented and from that information it can be
concluded that ACCT seems to lack the intangible resources needed to perform a digital
transformation. For example the ACCT project of incorporating mobile phones in the service
technician's’ daily work had a long execution time due to a lack of digital knowledge and
incentives. In order to perform this step successfully, ACCT needs to be aware of what they
are possessing. For this step, it is equally important as for the previous that ACCT obtain an
objective viewpoint when compiling company assets to avoid the risk of becoming outdated
by being overly self-confident.
Future visualization
As the third step in ACCT’s action plan, sustainable goals needs to be established. As
discussed in section 1, Introduction, the purpose of the current digital transformation is to
increase the company’s efficiency and profitability. Initially, to clarify these goals, these
KPI’s should be quantified. In addition to this, ACCT should establish long term goals that
are in line with the company's prevailing KPI’s. The purpose of not treating the digital goals
separately from the company's overall goals is to avoid misalignment and thus facilitate
prosperity.
Libert et al. (2016) claim that one should look at the transformation as a creation of a
digitalized network where the primary actor, in this case ACCT, can benefit from the external
actors. Consequently, ACCT should think outside their current barriers and embrace the
external environment. Tentatively, as mentioned in the above analysis of partnership, ACCT
should explore the opportunities of partnering with actors that possess knowledge that ACCT
lack. As the interviewees at ACCT described their core competences it was clear that IT and
digital technology was not included (see section 5, Empirical results) and it is therefore
recommended, as suggested in the three business models, that ACCT engage in alliances
where such knowledge can be gained.
Begin the transformation.
Wessel (2017) discusses the difficulty of getting all different incentives at a company to
become aligned. There will always be employees that disagree, as in the ACCT case
interpreted from the internal empirical results. In order to meet the uncertainties that a
transformation generates, Libert et al. (2016) recommend actors to conduct a pilot study while
Wessel (2017) argues for creation of small clusters. In line with these arguments and the
above analysis, pilot studies with an associated feedback loop are recommended to decrease
the future implementation time. In addition, pilot studies performed in small clusters with
visible results could facilitate the conviction of the more resistant employees. As was shown
in the empirical results (section 5), the digital resistance appear to be highest amongst the
service technicians, which constitutes a risk when implementing the business model that is
previously referred to as “Proposal 2” that implies service technicians using advanced digital
equipment. For this reason this area requires extra attention and it is recommended that ACCT
perform a pilot study at a department that appear more willing to accept new technology to,
when showing examples of successful implementation and the benefits of the new
technologies, gain trust from the more suspicious individuals.
57 |
Chalmers University of Technology | 7. Conclusion
In the following section, the conclusion of the thesis will be presented. By answering the three
research questions specified in section 1, Introduction, the following purpose of the thesis will
be achieved:
“The purpose of this thesis is to analyze how Atlas Copco Compressor Technique can
innovate its business model in order to increase efficiency and profitability, given
digitalization.”
RQ1: How could the Swedish air-compression industry be affected by digitalization
within the next five years?
Concluding from the findings, the compressor industry is developing and moving towards
increased digitalization. The main industrial implication related to digitalization that is
recognized is the prevailing threat of new entrants. The threat may appear by actors doing the
same things as incumbents, but in an improved way, or by disruptive actors substituting the
incumbent firms. At the same time, as companies embrace the new possibilities and
opportunities of digitization, the risk of companies disrupting the mature compressor industry
is evident. ACCT´s competitors are, similar to ACCT, characterized as incumbent firms
which indicate that they experience the same difficulties as ACCT in adapting to
digitalization. New entrants do not experience path dependency to the same extent as
incumbent firms and are thus often free to explore problems from different points of view
than the traditional. Highly digitalized new entrants, for which digital technology is a natural
part of the business, may one day solve the same problems for customers as ACCT does but
in a more innovative and efficient way. These new entrants may also use new innovative
business models that enables them to fully utilize the benefits of digitalization and thus
constitutes for the greatest threat for ACCT.
Actors within the compressor industry are implementing digital technologies aligned with the
concept Industrie 4.0 to increase competitiveness against new threats such as new entrants and
substitutes. Adapting to digitization has enabled competitors to alter their business models,
moving from product focus towards embracing a more holistic view of the purpose of their
business. This imply that within five years the industry will constitute of actors doing
business in alternative ways than today. Incumbents in the compressor industry are
developing their value propositions incrementally and are likely to follow in the footsteps of
the automotive companies and develop more digital products, still maintaining the basic way
of doing business with some exceptions. Incumbent clients are offered high quality products,
such as energy efficiency and predictive maintenance services.
Summarizing this section, the Swedish air compression industry could be affected by
digitalization in different ways. The Swedish air compression industry is likely to be
characterized by a larger amount of actors within the next five years due to the fact that
digitalization is making it easier for new actors to enter the market. The actors have likely
embraced digitalization and the use of differentiated business models while incumbent firms
still struggle with digital and innovative solutions.
RQ2: What opportunities and threats will digitalization imply for Atlas Copco
Compressor Technique in terms of products, processes and people?
The opportunities that emerge from digitalization are concluded to be energy efficiency and
value adding services and enhanced customer relationships. At the same time, these
59 |
Chalmers University of Technology | opportunities also appear for competitors which constitute as a threat to ACCT. The external
findings indicate that there are small differences between compressor suppliers’ competitive
offers, so it will be how well their offers are executed that will differ.
In order to capture the opportunities it is concluded that the transformation must come from
within ACCT. In establishing the processes and people needed to execute the digital
competitive strategy and thereby capture the product related opportunities, new ones arise. By
obtaining an engaged workforce and supporting processes, opportunities for further digital
development appear organically. Opportunities for digital development, by internal initiatives,
is captured by a shared vision and leadership combined with both top down and bottom up
information flows. Barriers towards such development are lack of knowledge and lack of
incentives to adopt new ways of doing things. It is recommended to encounter these threats, to
provide an organizational structure that enables two way communication, educating
employees and acquiring digital competence.
The main conclusion of this research question is that the product, process and people
dimensions are interconnected and mutually reinforcing. Opportunities created in one
dimension will enable new opportunities in the other two as well. Barriers towards those
opportunities are lack of knowledge and opposition towards doing things in a new way.
RQ3: How should Atlas Copco develop their business model within five years by
employing digitalization in order to expand their business through new revenue
streams?
Starting from today’s situation and the potential opportunities and threats that ACCT could
face in a transformation, three business models that can be considered interrelated have been
created. As the complexity of the problem lies in a lack of knowledge of how to respond to
digital change, the recommendation to ACCT is to adopt the three following business models;
one that offer energy efficiency solutions, one that offer more time efficient services at the
right time and one that offer provocative sales management in terms of offering the right
product when the customers are in need of a new one. In addition to these business models, an
action plan of how to embrace digital changes is provided as a complementary. The action
plan involves the following five steps; a clarification of the company’s starting point, an
identification of the company’s assets both tangible and intangible, a visualization of the
future in terms of establishing long term goals, an alignment of the different incentives within
the company and finally, measure the digital transformation process.
It can be concluded that by adopting the three business models presented, a sustainable
transformation can be achieved. The business models combined include digital elements that
can emphasize internal adoption processes and value creation in the form of efficiency and
customization. By combining technology with digital adapted internal processes and a
knowledgeable workforce, ACCT has the opportunity to develop their business through new
revenue streams. In addition, the recommended transformation will differ in execution time
between industry segments. Different industry segments require different approaches and
adaptation techniques whereby a recommendation will be to begin the transformation with the
consumer group early adopters. Subsequently, the remaining groups of customers can be
targeted. In order for ACCT to transform throughout the organization and to perform a
successful business model transformation, the emphasis must lie at the technological adoption
rate where the human perspective is the center of gravity (transformation).
60 |
Chalmers University of Technology | 8. Future research
In the following chapter, possible areas for future research will be presented and discussed
and above that, criticism of the report will be ascertained.
In hindsight, it can be argued that the result of the study can be applicable in other industries
as well. The mindset to transform the entire organization is equally important, regardless of
what sector the company operates within. However, it is essential to be aware of the possible
criticism that can be directed towards the study. Even if the study aimed to disarm the concept
of digitalization, it is important to be aware the risks that are involved. Companies must
actively decide on the level of transparency in order to avoid possibly fatal security breaches.
Thereby it is recommended to further investigate the possible risks a digital transformation
could entail.
This study can be seen as an indicator for ACCT to value the perspective of customers in a
larger extent. A future recommendation to ACCT is therefore to perform a more extensive
customer research with a wider range of participating customers, both in terms of number of
interviewees but also additional sectors and a wider geographical distribution. A larger sample
size could enable a higher level of credibility of the study and thereby result in a more
representable research to be used for further developments.
However, it can be discussed from another point of view, whether an extended customer
research would help or hinder a future digital transformation of ACCT. Both the focal
company and its customers are defined as incumbent firms and one of their common
denominators is the resistance to change and digital initiatives which leads to the question
how useful that type of research would possible be. As Henry Ford once said, “If I had asked
people what they wanted, they would have said faster horses”. If ACCT would like to keep
their front position in the line of compressor suppliers, they should consider the meaning of
Henry Ford’s quote and act from there. An alternative future recommendation for research
could be to, instead of asking the customers what they want, observe what they need. In a
world of constant change, thinking outside the box is essential for every actor, regardless of
the characteristic of the company.
Further related to customers, it would be of interest to study a new segmentation and
investigate its feasibility to ACCT. As observations indicate, customers from different sectors
have different needs whereupon they can be divided into differently segments. Since the
characteristics differs between these groups, they could be approached differently. This study
can be seen as an indicator for a possible change, however, to get an applicable segmentation,
a more extensive research should be performed before any decisions are taken.
62 |
Chalmers University of Technology | Abstract
Flotation is a very useful process. It is very popular due to its ability to collect small
particles from mineral slurry, for example reducing the grade of silica in iron ore; to
separate two kinds of mineral from each other. As more complex ores have to be
mined and higher-grade iron ore concentrates with lower impurities are requested
by the steel industry, achieving lower grades of silica in iron ore is a constant
struggle.
The aim of this work was to investigate the influence of different parameters on the
froth stability of silica flotation where silica is separated from magnetite. Magnetite is
a ferromagnetic mineral; its chemical formula is Fe O . The high-grade concentrates
3 4
of magnetite are used for iron ore pellet production in steel industry and for
advanced metallurgical applications.
Many parameters can affect the froth stability in mineral flotation. For example, type
of collector, concentration of collector, particle size, electrolytes and concentration of
electrolytes, airflow and stirring rate. In this work, the focus has been the effect of
stirring rate, airflow and concentration of collector.
There are a few factors used in froth stability evaluating, for example, froth maximum
height in equilibrium, growth velocity, decay velocity, froth retention time, air
recovery, froth half life time, etc. In this work, except froth half life time, the rest of
other five factors are compared to have a whole picture of the froth stability with the
change of different parameters.
The airflow plays an important role for froth stability when stirring rate is constant.
However, the stirring rate has major effect on the froth stability only when the
airflow is high. The alternation of the collector concentration can significantly impact
the froth height when the stirring rate and airflow are constant, nevertheless the
growth velocity is not affected by the collector concentration change.
1 |
Chalmers University of Technology | 1. Introduction
Froth flotation, which is over 100 years old as a patented process is widely used in
mineral processing plants. Nowadays, froth structure and stability attract more and
more attention because its significant role in determining the mineral grade and
recovery achieved from a flotation operation.
The aim of this master thesis was to investigate the influence of different parameters
on the froth stability of silica flotation where silica is separated from magnetite. The
aim of this project is to further develop the froth measuring method and to build up
understanding of how different parameters such as airflow, agitation and amount of
surfactant affect the characteristics of the froth.
A few factors were used to evaluate froth stability in this work, froth maximum
height in equilibrium, growth velocity, decay velocity, froth retention time, air
recovery. These factors can help to get integrate understanding of the froth.
This work has been divided in five parts:
Study the influence of different stirring rate on the froth behavior at constant
airflow.
Study the effect of different airflow on the froth behavior at constant stirring
rate.
Study the impact of different collector concentration on the froth behavior at
constant stirring rate and airflow.
Reproducibility of the test
MODDE-design of the two parameters: stirring rate and airflow
3 |
Chalmers University of Technology | 2. Theoretical background and literature review
2.1 Froth Flotation
Froth flotation, which was commercially introduced in mineral industry early in the
20th century about 100 years ago, is applied in many instances in the primary
mineral industries for separations of solids. Besides mineral industries, flotation can
also be used in wineries, food factories, smelters and refineries. After mining the
material from any deposit on the earth, crushing and grinding operations are applied
to liberate the mineral species. Processing water is added to produce the sample
insuspension, which is then added in the cell. Bubbles are induced into the slurry.
Surfactant acts in the interface of gas and liquid, and change the interfacial
properties. Modifying agent can be added in to the slurry depend on the kind of the
ore sample floated. The hydrophobic particles attach to the bubbles and are carried
to the surface of the slurry where the froth is formed. The surfactant is used to make
certain mineral particles hydrophobic, and leaving others in a hydrophilic form. [1]
2.1.1 Role of surfactant
Various particles with different surface properties can be separated by flotation
process. In this process, hydrophobic particles or hydrophilic particles which are
changed into hydrophobic by surfactants, attach to the gas bubbles in the pulp and
are floated to the froth, with most other particles remaining in the bulk. By
separating the froth from the bulk pulp, the hydrophobic particles thus can be
separated from the slurry.
Flotation reagents can control or modify the surface properties of minerals
selectively. By developing the flotation reagents, significant advances in flotation can
be made. Flotation depends on the probability of attachment of the bubbles to the
particles. The ability of the gas bubbles to attach to the particles is a function of the
hydrophobicity of the particles. When a solid particle is attached to a gas bubble in a
liquid, the resulting contact angle is given by Young’s equation in terms of the
interfacial energies
γ -γ =γ cosθ
sg sl lg
where γ , γ , and γ are solid-gas, solid-liquid, and liquid-gas interfacial tensions,
sg sl lg
respectively, and θ is the contact angle. Adhesion of the gas bubbles to the particles
in the suspension requires a contact angle that is significantly larger than zero. The
larger the contact angle, the greater is the hydrophobicity of the particle, and the
easier it is for the bubble to attach. [3]
4 |
Chalmers University of Technology | Figure 1: Contact angle between bubble and particle in an aqueous medium [4]
2.1.1.1 Cationic collectors
Because of the uniqueness of every ore, the reagents require careful test work to be
determined in each case. However, based on the prior similar test work, it might be
possible to predict the suitable reagent for a new composition of ore.
There are a few minerals, such as natural sulfur and graphite, which are naturally
hydrophobic. In most cases, the mineral particles are hydrophilic and surface-active
agents are added to impart the hydrophobicity to the particles. [3]
Both anionic and cationic collectors can be used for this purpose depending on the
charge of particles. The cationic collector will be emphasized below.
In cationic collectors, the nitrogen atom carries the cationic charge. Common
products are amine and quaternary ammonium-based products [2]. These types of
collectors have been in industrial use since the mid 1930’s and are especially useful
in the flotation of silicate minerals.
The general structure is:
R-NH
2
NH2 is the polar hydrophilic group and R is the hydrophobic tail, hydrophobic carbon
chain comprises 8-24 carbon atoms. The longer the carbon chain is, the harder the
amine is to disperse.
These molecules will ionize in solution. The ionized molecule has a positively charged
head that will attach on the silicates’ negatively charged surfaces, and the
hydrophobic tail will attach to the air bubbles [14].
Since pH will change the polarity, cationic collectors must be used in carefully
controlled pH. The amines are weak bases and occur as cationic ammonium ion at pH
below 10.
The change in the hydrophobicity of the adsorbed layer, and not the occurrence of
5 |
Chalmers University of Technology | the adsorption bond itself influences the floatability of mineral particles. It is known
that an increase in the alkyl chain of a given collector improves floatability and
lowers the threshold concentration needed for flotation.[1]
the formation of ion-molecular complex at a certain pH range is an important
property of amine collectors.
2.1.1.2 Critical micelle concentration, CMC
When measuring the different physicochemical properties of an aqueous solution of
a surfactant, many peculiarities will be encounter, for instance cationic surfactants, as
exemplified in figure 2.
Figure 2: Schematic representation of the concentration dependence of some physical properties
for solutions of a micelle-forming surfactant[2]
From the above figure, at low concentrations, most properties are similar to those of
a simple electrolyte. But surface tension shows distinct exception. Giving a notable
physical property to be measured to get CMC of the surfactant, since it decreases
rapidly with surfactant concentration,.
Along with the increasing concentration of the surfactants, they tend to assembly.
The first-formed aggregates are generally approximately spherical in shape. Such
aggregates are called micelles, and the concentration where they start to form is
known as the critical micelle concentration, abbreviated to CMC.
The CMC will change with type and concentration of electrolytes and of course with
changes in pH.[2]
2.1.2 Role of electrolyte
There are many factors that can influence plant flotation. The presence of interfering
ions in the process water is one of the most common one. These ions may come
from many different places in the process, for example it could be organic chemicals
6 |
Chalmers University of Technology | such as tannic acid and lignins or it can come from the water added to the process.
The process water is recirculated and will be almost saturated with dissolved ions
from the ore.
Multivalent ions such as calcium and magnesium can be particularly devastating for
amine flotation [15]. Together with the increase of concentration of magnesium and
calcium, an increase in collector consumption and loss of selectivity will take place,
because of the increased competition for negative sites on the surface of the mineral
particles [16].
pH will dramatically influence the effect of these ions on the flotation. The recoveries
are affected very little by calcium or magnesium at pH 7. The recoveries gradually
decrease over pH 10.5. An explanation to this could be the absorption of the ions on
the mineral. Adsorption measurements have shown that the calcium ion is adsorbed
as hydrated Ca2+ at pH 7, whereas it is adsorbed as CaOH+at pH 11. [17].
These include the action of the electrolytes in disruption of hydration layers
surrounding the particles and enhancing bubble-particle capture, reduction of the
electrostatic interactions, and an increasing charge on the surface of the bubbles to
prevent primary bubble coalescence.
The flotation efficiency will increase with additional salt concentration.
The inhibition of bubbles to coalescence in electrolytes can be realized by the
application of a combining rule based on the nature of the cationic/anionic pair.
Viscosity and electrostatic repulsion were ruled out as possible explanations. In fact,
following conventional electrostatic double-layer theory, and increase in salt
concentration would reduce the double-layer repulsion and should induce inhibition
[1].
2.1.3 Role of pH
In flotation, the pH of the slurry is crucial, as in solution, the extent of ionization,
hydrolysis of surfactants and the charge on the mineral surface is determined by pH.
This in turn influences the possibility of the collector to attach on the surface.
Therefore, at the various ionized solid/liquid interfaces, pH can either helps or
hinders the adsorption of the surfactant, contributing to greater or lesser selectivity
of flotation [1].
Researchers have established that at an approximate pH of 2, silica has zero charge,
amphiboles at 5 and magnetite at 7. This phenomenon explains why at a pH of 8-9,
silica floats more easily, amphibole floating next and magnetite only when excess of
reagents is added to the pulp.
Flotation usually takes place in the pH range of 8 to 11[18]. It is not favorable to run
the process with an acid pH since this will damage the equipment due to the high
7 |
Chalmers University of Technology | tendency of corrosion and dissolve iron oxide.
2.1.4 The major forces in surfactant adsorption
Electrostatic adsorption of surfactants
Electrostatic interactions play a governing role in the mineral-surfactant adsorption
process in the system where the surfactants and the particles are oppositely charged.
Charge on the particle surface, in the case of oxides, is a result of the hydrolysis of
surface species followed by pH-dependent dissociation of the surface hydroxyl
groups:
MOH = MO- + H+
H+ + MOH = MOH +
2
where M represents the interfacial metal atom. The pH at which the surface charge is
zero is called the point of zero charge (PZC). In the solutions which are more acidic
than the PZC, the oxides carry a positive charge and a negative charge in those is
more alkaline [3]. The adsorption of amine-silica follows this mechanism. When the
solution is more alkaline than PZC, in this case pH=8, the surface charge of silica is
negative, attracts the positive charged etherdiamine collector.
Chemical adsorption of surfactants
Between the surface site of the particles and the polar head group of the surfactant,
chemical adsorption or “chemisorption” takes place through covalent or coordinate
bonds. The result is the formation of a new surface compound. Chemisorption takes
place in the systems where metal ions don’t leave their lattice sites. This kind of
chemisorptions is termed “surface reaction”, when metal ions composing the mineral
move out of their lattice sites and interact with the collector,.
If chemisorption took place in the system, it would be the dominant froth compared
with electrostatic adsorption of surfactants.
Hydrogen Bonding
For adsorption of amount of reagents, particularly those containing hydroxyl,
phenolic, carboxylic and amine groups, hydrogen bonding between the surfactants
and the particle surface is accountable.
2.1.5 Hydrodynamics of mineral flotation
The collision between particles and bubbles in the liquid suspension, attachment of
the particles to the bubbles, and levitation of the particle-bubble aggregates to the
surface of the suspension assume vital roles in flotation in addition the forces
between the surfactant and the particle surfaces.
8 |
Chalmers University of Technology | Film rupture between minerals and bubbles
In determining the stability of bubbles and droplets in froth, rupture of thin liquid
films between liquids, gases, and solids plays a governing role.
Particle-Bubble interactions in Solutions
Many varieties of forces can result in attachment of particles to bubbles, such as
electrical double-layer forces between the particle and the bubble (with adsorbed
surfactant on both), Van Der Waals forces, the energy change due to the transfer of
the hydrocarbon chains adsorbed on the particles to the gaseous phase.
Electrical Double-Layer Interactions
In the mechanism of the electrostatic stabilization of colloids, electric Double Layer is
the phenomenon playing a fundamental role. When negatively charged ions of the
dispersion medium are adsorbed on particles surface, colloidal particles get negative
electric charge. A negatively charged particle attracts the positive counterions
surrounding the particle. Electric Double Layer is the layer surrounding a particle of
the dispersed phase and including the ions adsorbed on the particle surface and a
film of the countercharged dispersion medium. As shown in figure 3.
Zeta potential is electric potential in the interfacial double layer (DL) at the location
of the slipping plane versus a point in the bulk fluid away from the interface. In other
words, zeta potential is the potential difference between the dispersion medium and
the stationary layer of fluid attached to the dispersed particle as in figure 3. [3]
Figure 3: Electric double layer and Zeta potential [20]
Zeta potential varies with the distance from the particle as it is shown in figure 3; The
zeta potential decreases when the distance to the surface increase. When two
9 |
Chalmers University of Technology | particles move closer to each other, the overlap of double layers will lead to the
interaction energy change.
Van der Waals interactions
When the bubble and particle move together, an energy change will be associated
with van der Waals forces, because the present system is comprised of three
different media (solid, air, and intervening solution).
Interaction due to Steric Repulsion
When the interacting spheres (bubble and particle) contain adsorbed layers, their
adhesion will be sterically hindered, mainly because of the physical size of the
molecules.
2.2 Two phase foam and three phase froth
The froth phase in column flotation plays a very important role in determining the
mineral recovery and the grade of produced concentrate. The role of the froth has
only relatively recently been recognized as important, and as being more than a
simple mass transfer mechanism for carrying particles attached to bubbles from the
pulp- froth interface to the concentrate launder.
2.2.1 Two phase foam
Two phase foam is a dispersion of a gas in a liquid or solid. Here only liquid foam will
be elaborated. [2] In the foam structure, the bubbles are separated by a thin water
film, called lamellae, and the intersection of three lamellae is called Plateau.
Figure 4: Bubbles consist of thin films (lamellae) meeting in lines (Plateau borders)[6]
10 |
Chalmers University of Technology | 2）the particles themselves are surface active. By adsorbing, or chemically reacting,
hydrophobic moieties on the surface of the particles, hydrophilic particles can be
changed into hydrophobic. Too high hydrophobicity of the particles will precipitate
the particles and destroy the foam. If the particles are made only partly hydrophobic,
however, they will show surface activity, i.e. they will have an affinity for the liquid
and air interface. Such systems give very stable foams. When the contact angle of
water is around 90°, the best foam stability is achieved, i.e. when one half of the
particle is in the liquid and the other half is in the air, as shown in Figure 7.
Figure 7: Solid particles act as foaming agents if they are made partly hydrophobic. The best foam
stability is achieved when the contact angle is around 90°[2]
Table 1: a summary of the four forces acting on foam[2]
2.2.2 Three phase Froth
In the flotation process it is clearly recognized that there are two distinct particulate
systems: a pulp phase and a froth phase. [9]
Froth is a complex three phases system consisting of air bubbles, particles and water,
its ability to transport particles into the concentrate launder is crucially determined
by the structure and stability of the froth. Flotation froths reach a stable height in the
column after a period of time, however, two-phase foams which do not reach an
equilibrium height but continue to grow indefinitely. [9] Froth structure (including
the bubble size distribution, solids loading and liquid content) are known to play a
12 |
Chalmers University of Technology | significant role in determining the grade and recovery achieved from a flotation
process.[10]
2.3 Froth test
A variety of methods has been developed to assess the stability of froth. In general,
two types of tests can be identified: dynamic tests and static tests.
A dynamic test is one in which the foam has reached a state of dynamic equilibrium
between rates of formation and decay.
A static test is one where the rate of foam formation is zero: the foam, once formed,
is allowed to collapse without regeneration by further agitation or input of gas.
Dynamic test comes closest to the dynamics of a real flotation system. The air is
continuously introduced into the solution, the foam starts forming at the bottom and
then moves upwards. The liquid drains downwards so that quality of the foam varies
with height. The top of the foam then reaches the breaking condition and the top
layer collapses. Those features are similar to the real flotation and make the dynamic
method more likely to give a true indication of the froth stability during a flotation
test. [5]
2.4 Influencing parameter of froth stability
The surfactants play a major role in foam stability. They adsorb at the gas-liquid
interface and change the interfacial properties. The influence of surfactants on foam
behavior is not clear. But it is worth noted that froth stability is mainly controlled by
particles, rather than surfactant molecules.[5]
2.4.1 Motion of the particles
For both hydrophobic or hydrophilic particles in the froth, two possible types are
considered: those particles attached to the bubble lamellae or those particles that
are unattached and free to move through the Plateau borders. The attached particles
will mainly only be hydrophobic, whereas the unattached particles may be
hydrophobic or hydrophilic.
In most cases, the bubbles are significantly larger, often as much as two orders of
magnitude, than the particles which are attached to the bubble lamellae.
The Attached Particles
Because of bubbles bursting at the top surface or through coalescence in the froth,
attached hydrophobic particles detach from bubble lamellae and become unattached.
Some of the bubble surface area is lost, and the attached particles on that lamella
transfers to the Plateau borders, when bubbles coalesce or rupture. The role of the
unattached particles is often much more significant than that of the attached,
because there is a tremendous number of coalescence occurring in the froth, and
13 |
Chalmers University of Technology | also a significant fraction of the air bursting on the surface.
The Unattached Particles
The unattached particles move through the Plateau border network and will tend to
follow the liquid.
2.4.2 Nature of the particles
The nature and concentration of the particles present in the system can also
influence the froth stability. Three important parameters are 1. the capability to
stabilize the liquid films separating the bubbles; 2. the facility to form a
three-dimensional network in the bulk aqueous phase; 3. the ability of the particles
to form coherent shell around the bubbles.
Particle size
The dynamic froth stability depends on the particle size. Previous work shows that
equilibrium height increase when particles become finer, and hereinafter, the
dynamic froth stability increase. Particles can arrange at the liquid/gas interface and
stabilize the films by capillary mechanism, if the particle size is small compared to the
film thickness. However, particles can bridge and rupture the foam films if they are
large. It has been reported that as the particle size decreases, when using <0.7um
silica particles, froth become more stable.
Particle shape
The shape of the particles is also critical. The thinning and rupture of liquid film can
be accomplished in roughly 0.1s by rounded or spherical particles, however
sharp-edged particles rupture the liquid film in about 0.02s. [5]
Hydrophobicity of the particles
When hydrophilic particles are present, the stability of particle-stabilized foams is the
lowest. Dippenaar and Harris discovered that very hydrophilic particles of all sizes
can destroy froth, and the destabilizing effect is independent of the frother type,
unless the frother adsorbs onto the particles. Similar effect would appear, when it
comes to the particles of high hydrophobicity. These particles may break films and
suppress the froth stability, whereas those with lower hydrophobicity have little
effect on the froth stability because they remain dispersed in lamellae.[3]
Gourram-Badri (1997) found that the presence of particles in the foam determine
the total foam height, the total foam height is lowered with the increased amount of
particles, moreover they found that more hydrophobic particles do not produce a
stable steady-state height.[7]
2.4.3 Bubble size
Bubbles, having hydrophobic particle load associated with them, enter the froth from
14 |
Chalmers University of Technology | the pulp/froth interface. The bubble size, density, and coverage of the particles
determine the particle load that a bubble can carry. The particles are found on the
lower hemisphere of the bubbles, and are free to move on the surface of the bubble,
rising through the pulp.
Both by the length and volume of Plateau borders containing the unattached
particles and by the lamellae surface area containing, the attached particles The
bubble size can affects the flotation process dramatically. The lamellae size
distribution is not as the same as the bubble size distribution, and will depend not
only on the bubble size, but also the shape.
2.4.4 Bubble coalescence
The rupture of thin liquid films that separates two adjacent bubbles is one of the
main reasons for bubble coalescence in froth. [6] When the liquid content in the
froth falls below a critical value, this occurs. The presence of particles and surfactants
stabilize the bubbles, therefore they are believed to have a detrimental effect on the
film rupture.
Due to the presence of hydrophobic particles, bubble coalescence in flotation froths
will also occur. A spherical particle will attempt to move toward a central position to
meet the contact angle requirement, when it enters a foam film and bridges the two
interfaces. A stable orientation will be attained and the life of the film will be
prolonged, if the contact angle is less than a critical degree of wetting. The particle
will easily dewet through both sides of the lamella, resulting in the film rupture as
shown in the fig below, in the case where the bridging particle has a contact angle
more than the critical degree of wetting.
A) Contact angle of surfactant solution B) Contact angle of surfactant solution
with particle< 90°.with particle > 90°.
Figure 8: Bridging particle behavior in a foam. A) Moderately hydrophobic, B) highly hydrophobic
particle [6]
15 |
Chalmers University of Technology | 2.4.5 Detachment of hydrophobic particles
Hydrophobic particles detach from the surface of bubbles within flotation froths. The
detached particles may be recaptured and transported into the froth when they
return to the liquid zone. There are several reasons of the detachment of floatable
particles from the froth: Coalescence of bubble is one of the primary reasons. In
flotation froths especially at the base and surface of the froth, bubble coalescence
occurs constantly. A high rate of coalescence may eventually lead to froth
overloading (maximum carrying capacity) to a point where further transportation of
floatable particles to the concentrate may not be possible.
In a relatively heavily loaded froth or at the top of the froth where the surfaces of the
bubbles are fully covered, the detachment due to insufficient surface area is
expected to be effective. At the pulp-froth interface where bubbles rising from the
liquid beneath start to crowd and pack together, it is also largely occurs.
During the merging process as a result of the oscillations caused by the coalescence,
the coalescence of two bubbles could be vigorous and that the particles residing on
the surface of the bubbles could be detached. With adsorbed surfactant used to
increase contact angle and thus reduce particle detachment, particle hydrophobicity
played an important role. [6]
2.4.6 Frother concentration
The bubble coalescence can be decreased by increasing frother concentration, and
the coalescence of the bubbles is prevented at a particular concentration, known as
critical coalescence concentration (CCC). As it is shown in the schematic figure 9,
coalescence does not occur at concentrations higher than CCC.
Figure 9: Effect of frother concentration on bubble size (schematic)
The effect of flotation frothers on bubble size and its implications for foam stability
were studied by Cho and Laskowski (2002). The CCC values and the dynamic
foamability index were correlated, and that is determined by bubble coalescence.
Foam stability under dynamic conditions is determined by bubble coalescence. [7]
16 |
Chalmers University of Technology | 2.4.7 Operating parameters
Conditioning time
Conditioning of flotation feed particles with various chemical reagents is an
important parameter. The maximum froth height declines as the conditioning time
increases, it might be because the particles hydrophobicity increase and exceed the
critical value for stabilizing the froth.
Stirring rate
In froth test, mechanic stir is needed in the cell in order to break down the bubbles
which are produced from the nozzle. This in turn will lead to larger surface area in
total and make the particle attach to the bubbles and float up more easily.
Airflow rate
Airflow rate is of great importance in determining the flotation rate constant. It can
influence the superficial gas velocity in the froth, which is defined as the gas blew in
the cell and that grows up into the column as bubbles regardless of the bubble
collapse and coalescence.
Froth stability to flotation performance with a series of coal column flotation
experiments was directly related by Tao (2000). Airflow, froth height, and feed solids
concentration rate was manipulated, it is found that an increase in the airflow rate
stabilizes the froth, however operating at superficial gas velocities higher than 2 cm/s
results in a significant increase in the ash recovery due to non-selective entrainment.
[11]
2. 5 Analyzing factor for froth stability
Used for separating mineral ores, the froth flotation process depends on the
establishment of a carefully controlled and stable froth. Measuring the froth height
for a given frother dosage and a given airflow rate in a measuring cylinder is a simple,
semi-quantitative method of estimating the frothing power of a frother. The
relationship between flotation variables, performance and froth structure was
determined by previous work. It has become clear that froth stability plays a major
role in flotation performance, from these studies. Nevertheless, there is no unique
measure, or precise definition, of froth stability as yet. Many different methods to
measure froth stability have been employed over the years.
Froth maximum height at equilibrium
From the pictures which are taken from the froth test, the froth maximum height at
equilibrium is measured directly. The froth height H is measured as a function of time,
t, using the pulp-froth interface as the reference when the froth rises up inside the
flotation cell and grows upwards in the measuring column. The froth rise stops and it
reaches a constant height, H , after some time, depending on the froth stability
max
17 |
Chalmers University of Technology | and operating conditions. [12]
The rate of froth growth and the maximum equilibrium height change for different
airflow rates and frother concentrations was studied by Barbian(2003). In this study,
when increasing airflow rate and frother concentration, the maximum equilibrium
froth height was found to increase. However, the maximum equilibrium height
decreases, at high aeration rates and frother concentration. Under these conditions
the froth is no longer stable and collapse. [7]
Froth growth velocity
Froth growth velocity is the rate or velocity of movement of froth up the column
from a pre-determined starting height to a maximum height of the froth in the
column. A patent “Measuring froth stability”measures the froth height in a cell of a
flotation circuit for slurry and plots the data over time. They used a fitted model
t
−
H = Ho 1−e Г to fit the curve and their data is described perfectly by this model,
as shown in the fig below.
Figure 10: Column forth height as a function of time
They also introduced plenty of other parameters to make the most use out of the
model. For instance,
The superficial gas velocity Jg
volume flow of gas
which is defined as Jg =
cross sectional area of the cell
It is the gas flow rate per unit area of the column, this parameter is calculated based
on the assumption that all the air entering the froth from the pulp was retained in
18 |
Chalmers University of Technology | represents (the average time that gas remains entrained in the foam) lifetime of a
bubble in the froth, and is defined as the ratio of the total volume of froth at
equilibrium to the volumetric gas rate introduced into the system. It is expressed as
V H A
f max
Σ = =
Q Q
Where V is the foam volume, Q is the gas volumetric flow rate and H and A are
f max
the total foam height and the cross-sectional area of the vessel, respectively.
The experiments at laboratory scale to measure the dynamic stability factor and
equilibrium volume of the froth under different operating conditions—different
airflow and frother concentration was carried out by N.Barbian (2003). It was found
that the dynamic stability factor decreases as the airflow rate increases. At low
frother concentration, Σ is lower than at higher frother concentration, however as
the airflow rate is increased the value of Σ rapidly decreases. This implied that the
operating variables can significantly influence froth stability; Σ is affected by airflow
rate and frother concentration. Together with the result which was achieved from
the froth growth and the maximum equilibrium height measurement, it was also
found that dynamic stability factor were used to predict the expected fraction of air
overflowing below the maximum froth height.[7]
It is necessary to note that in some articles, the Froth retention time (FRT) is used to
describe the average lifetime of an air bubble in the froth system, from mathematic
point of view FRT is the same as Σ.
Hmax
Froth retention time defined as FRT= by Laplante (1983), where H is the
max
Jg
froth depth and Jg is the superficial gas velocity.
The froth retention time is physically a measure of the average lifetime of an air
bubble in the froth system. For a given froth the longer the froth retention time FRT
is, the greater the probability for an attached particle to be dropped back into the
slurry is. The increase of water drainage and reduce of entrainment, which is often
desirable, take place when the froth height is increased. On the other hand, the froth
recovery of the valuable minerals may also be reduced.
A more accurate definition of the mean froth retention time was given by Zheng
(2006), who consider that: a) not all of the volume of the froth is occupied by air, and
b) due to bubble bursting on the surface, only a fraction of the total airflow
introduced to the cell reports to the concentrate.[9]
Air recovery
Air recovery, which is the fraction of air, flew into a flotation cell that overflows the
cell lip as unburst bubbles. The air recovery is, in general, surprisingly low, and values
greater than 50% are rarely observed.
20 |
Chalmers University of Technology | Air recovery empirically determines the balance between the valuable particle
recovered by attachment to the lamellae and those recovered from the Plateau
borders. The air recovery determines, to a large extent, water recovery, which is
related to gangue recovery.
The air recovery, α, is the fraction of air overflowing the weir that entered the cell:
α= Q / Q
air, concentrate air, pulp
Q relates to the airflow rate which is set for the experiment,
air, pulp
Q = u * H
air, concentrate froth froth
Half life time t
1/2
Half life time is a characteristic time is used to describe froth stability. It is the time
taken for the foam to drop to one half of its initial height(Hmax) when the gas flow is
discontinued. In addition to the factor to determine the froth stability, it is also a
parameter for a model for decay, t together with α is crucial parameters to
1/2
H 1
determine froth decay. It is described in the study of Zanin(2008) as : = −α∗
Hmax 2
t
ln( ) , this model fitted the experimental data they gathered from a plant. [9]
t
1/2
2.6 Iron ore
2.6.1 Magnetite
Fe O , is exactly the same name says a magnetic ore. Magnetite is a member of the
3 4
spinel group. This group has the general formula AB O . A and B usually represent
2 4
different metal ions but in this case A is Fe2+ and B is Fe3+. This arrangement is called
an inverse spinel and causes a transfer of electrons between the different irons in a
structured way. It is this “electricity” that causes the magnetic field that is
characteristic for magnetite.
Magnetite is a common mineral found in most types of igneous rocks. The largest
magnetite deposits in the world are found in northern Sweden.
Magnetite is black with a metallic luster. The crystal system is isometric. The crystal
usually forms as octahedral shape.
The best way of recognizing this ore is by the magnetism, thereafter by its streak and
by its crystal habit.
2.6.2 Silica
Silica, also known as silicon dioxide, is an oxide of silicon with the chemical formula
SiO . Since antiquity, it has been known for its hardness. In nature, silica is most
2
21 |
Chalmers University of Technology | commonly found as sand or quartz. There are three stable forms of silica structure,
i.e., quartz (stable below 870℃), tridymite (870-1470℃), and cristobalite (above
1470℃). The Si-O bonds possess a considerable degree of covalent character, with
oxygen atoms arranged tetrahedrally around Si in such a manner that every oxygen
atom is shared by two SiO tetrahedra. [1]
4
2.7 Experimental design
The method will give an organized approach of a set of experiments and it helps to
see interactions between the factors are one of the biggest benefits with experiment
design. This is a very effective method to reach a better result, if there is more than
two variables, probably with fewer trials than there would have been otherwise.
Therefore experimental design is often used in optimization work where the
experimental cost otherwise would be very high.
General
There are three objectives in optimization work where experimental design could be
used: screening, optimization and robustness testing. Three main problems that
often occur in those kinds of works are coped with by the method:
How to monitor a systems where many factors simultaneously influence the
system
Separation of real effects and noise
Producing reliable maps
A good experimental design will give a set of experiments where all relevant factors
are varied simultaneously and this set of experiments should of course be
representative for the given question.
Create a standard representative experiment, a centre point, and then perform other
representative experiments that lie symmetrically around this standard is one way to
look at the experimental design. This gives a square pattern. If there are more than
three factors to change, it will be a more complex pattern. One of the most
important things with experimental design is that these trials should be performed
randomly. [19]
22 |
Chalmers University of Technology | 3. Experiments
This work was focused on the different parameters which influence the froth stability.
The experiment is launched by changing the stirring rate, airflow and the collector
concentration respectively with the other two parameters constant. It is important to
note that electrolyte is also one of the significant parameters for froth stability. In
this work the electrolyte remained the same in all tests.
3.1 Parameters
Process water:
For all mineral flotation, the electrolytes in the process water are different. In this
work electrolytes were added in the water to simulate the process water in a plant.
The table shows the concentrations under which experiments were carried out. All
salts are held high quality.
pH Ca Mg SO Cl HCO
4 3
mg/l mg/l mg/l mg/l mg/l
process water 8 600 50 1900 530 50
Table 2: Salt type and concentration of the process water
Collector:
An Etherdiamine is used in this work. The chemical structure of this Alkyl
etherdiamine is R-O-(CH2)3-NH-(CH2)3-NH2 with isotridecyl as alkyl chain.
Ore sample:
The iron ore used in this work is the magnetite, 56% Fe, 16 % acid insoluble (the
amount of silicates is measured as acid insoluble) Particle size: K80 = 34,5 µm (K80
means that 80% are smaller than 34,5µm).
3.2 Procedure:
General:
Every single test is run by the same procedures below with different parameters
setting. This froth test instrument is similar to many froth studies in laboratory scale.
The flotation cell with the mechanical stir at the bottom in the froth test is a rebuilt
open-top flotation machine.
Collector solution preparation: make 1.0 wt% solution
The solution is stirred with magnetic stirrer for 30 minutes in order to fully mix
the collector with water
Airflow regulation: set the airflow to the experimental request every time before
each test
Preparation of the flotation cell: add 644g ore and the process water up to 1.3l,
set the stir rate and stir the dispersion for 2 minutes before adding the certain
23 |
Chalmers University of Technology | amount of collector. Stir the dispersion for 2 minutes after the collector is added.
Build up the froth column and place the measuring ruler tape on the top of it as
the figure below.
Figure 12: Picture of the whole experimental set
The airflow and the camera, the camera is set to take a picture of the froth
column every 20 second.
Turn off the air when the froth has reached its first maximum height, wait a
couple of minutes to allow the froth collapse without any external disturbance.
The
Spray tap water on the top of the froth in order to get faster decay of the froth.
The water is sprayed approximately every 2 minutes, twice or three times
depends on the froth height of self decay.
Stop the test when the froth approximately decayed to the interface level
between pulp and bubbles
Data reading from the pictures taken by the camera
Specific:
1. Froth behavior under the change of stirring rate with 1l/min airflow, and collector
concentration of 144g/ton.
Stir set: 650rpm, 900rpm, 1050rpm and 1200rpm
2. Froth behavior under the change of stirring rate with 3l/min airflow, and collector
concentration of 144g/ton.
Stir set: 650rpm, 900rpm, 1050rpm and 1200rpm
3. Froth behavior under the change of airflow with 900rpm stirring rate, and
collector concentration of 144g/ton.
24 |
Chalmers University of Technology | Airflow set: 0.5l/min, 1l/min, 1.5l/min, 2l/min, 2.5l/min, 3l/min and 5l/min
4. Froth behavior under the change of collector concentration with 900rpm stirring
rate, and1l/min airflow.
Concentration set: 36g/ton, 72g/ton, 144g/ton and 288g/ton
5. Reducibility set: collector concentration 144g/ton
650rpm, 1l/min; 900rpm, 1l/min, 900rpm, 3l/min, 1200rpm, 1l/min and
1200rpm, 3l/min
It should be mentioned,
The dry weight of every ore sample, which is used in froth test, is 644g, and the
content of solid-in-pulp is 35.5wt%.
When running the experiment with constant collector concentration, 144g/ton
of collector is used as standard dosage in this work. This is because in the
flotation test, the target 4% acid insoluble in the magnetite concentrate is
reached at 144g/ton of collector. Acid insoluble is an indirect factor to show
gangue content in magnetite concentrate.
Because 900rpm has longer froth equilibrium stage, this stirring rate is chosen as
the standard stirring rate for the change of airflow rate and concentration.
3.3 Analysis:
Froth height
This is the most direct data we can get from the pictures taken by the camera, and is
also the y axis factor in the froth behavior diagram.
Froth growth velocity u (cm/s)
This is measured from the maximum slope in the froth growing part in the diagram.
Froth retention time (s)
Besides froth height and growth velocity, the froth retention time (FRT) is a factor
used in this work
Hmax
FRT=
Jg
FRT is calculated by knowing the froth height at equilibrium from the diagram.
Froth decay rate v (cm/s)
This factor is achieved from the fitting line fitted out from the point when the airflow
is turned off to the end of the test, during the time water is sprayed on the top of the
froth with the amount 11-12 g three times.
Air recovery α (%)
Air recovery has been recognized empirically as determining the balance between
the valuable particle recovered by attachment to the lamellae and those recovered
25 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.