University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Chalmers University of Technology | from the Plateau borders.
α= Q / Q =Jg/u
air, concentrate air, pulp
3.4 Experimental design
In this work a computer program called MODDE which is invented by Umetrics is
used to do the experimental design.
Design set up
To begin the design some input conditions have to be defined. There has to be a
number of different factors, for example stirring rate and airflow, and in which range
they should be varied or if they are uncontrolled. Also the number of responses and
their units has to be specified. The experimental data was putted into MODDE. Each
experiment will give a result (a response) that will be analyzed by regression and a
model will be given where the changes of factors will be related to the changes in
responses. This model will indicate which factors that is most important for the
optimization and how they should be combined to get the desired results.
In this work, the external influencing parameters for froth stability, which in this case
are four different stirring rates and seven different airflows, maximum height and
growth velocity, are studied by the help of MODDE.
Factors: Stirring rate (experimental set, controlled, 650, 900, 1050, 1200) [rpm]
Airflow (experimental set, controlled, 0.5, 1, 1.5, 2, 2.5, 3, 5) [l/min]
Response: Maximum height (cm)
Growth velocity (cm/sec)
Air recovery (%)
Decay velocity (cm/sec)
26 |
Chalmers University of Technology | The figure 17 shows the result from the influence of stirring rate 650 rpm, 900 rpm,
1050 rpm and 1200 rpm on the froth stability with the airflow of 1 l/min and
collector concentration 144g/ton. For all the stirring rates, they had almost linear
growth after 60 seconds up to the equilibrium height. 650 rpm and 900 rpm
respectively 1050 rpm and 1200 rpm have similar growth rate. However, the
difference between two rates is not so obvious. The equilibrium heights (maximum
height) are among the same level for all stirring rates. The froth of 1200 rpm didn’t
behave like others. It had an abrupt drop right after achieving the maximum height.
However for the stirring rates which are lower than 1200rpm, when the froth height
came to the equilibrium, they can remain in this equilibrium state for some time with
slight froth height fluctuations. The decay stage is quite difficult to interpret from this
diagram, because of the different time of the water spray, and the different bubble
coalescence for different stirring rate.
50
45
40
35
m 30
c
/
th 1200 rpm
g ie 25
h 900 rpm
h
to
rf 20 650 rpm
15
10
5
0
0 200 400 600 800 1000 1200 1400 1600
time/sec
Figure 18: Froth behavior with different stirring rate at 3 l/min airflow
The figure 18 shows the result from the influence of stirring rate 650 rpm, 900 rpm,
1050 rpm and 1200 rpm on the froth stability with the airflow of 3 l/min and
collector concentration 77 mg/l. As it seems in the figure for 3 l/min, the froth had a
linear growth before it came to the equilibrium height. The differences of growth
velocity and maximum height for the different stirring rates were quite obvious. The
growth velocity and the maximum height decrease as the stirring rate decreases.
Besides maximum froth height and growth velocity, there are other froth stability
factors which are also valuable for froth stability evaluation. Froth retention time
30 |
Chalmers University of Technology | 0.25
0.2
0.15
s
/
m
c
/ u 1l/min
0.1
3l/min
0.05
0
0 200 400 600 800 1000 1200 1400
stir rate/rpm
Figure 20: Variation in growth velocity for different stirring rate and airflow
Figure 19 and 20 are plotted out from the table 3 and table 4. It shows directly the
distribution and the variation trends of the stability factors between 1 l/min and 3
l/min. Figure 19 and 20 shows the height and growth velocity distribution for
different stirring rate at different airflow respectively. The reason that these two
factors at 3 l/min are higher than the ones at 1 l/min is: the higher the airflow is, the
larger the amount of bubbles which build up the froth are in the system. For 3 l/min,
these two factors are ascending as the increase of the stirring rate. This is because
the higher the stirring rate, the smaller the bubbles becomes. Consequently, there
are more surface area for the particles to load on, in turn the froth can be stabilized
more easily. From the distribution for the factors at 1 l/min, it is reasonable to
conclude the stirring rate is not the dominant parameter for the froth stability when
the airflow is low. This might be because the bubble amount is the crucial parameter
for particle loading when the air content is low. Even though increasing the stirring
rate can increase the total surface area of the bubbles, the amount of particles that
can load on to the bubble surface depends on the bubble amount instead of the
bubble surface area. It is acceptable to presume there is a critical airflow rate, which
is the watershed of these two influencing mechanism.
32 |
Chalmers University of Technology | 0.07
0.06
0.05
0.04
)s
/
m
c
(
v 0.03
0.02
0.01
0
0 1 2 3 4 5 6
air flow (l/min)
Figure 25: Variation in decay velocity with different airflow for 900 rpm
Figures 23 and 24 verify the results from figure 22. This phenomenon is explained by
the air content increase in the froth when the airflow rises up. More bubbles exit in
the system which leads to more hydrophobic particles attached to the lamellae, also
gangue particles reside in the plateau. This can make the froth more stable. However,
when the amount of bubbles grows up to a certain number, in this case at 3 l/min,
the growth velocity is not increasing any more. This is predicted by knowing that the
particles content is constant, even though higher airflow rate gives a higher amount
of bubbles and total surface area, the amount of particles is not enough to attach to
all the surface of the bubbles. The froth height comes to a certain level when the
airflow rate rises up to 2l/min. This is also due to the same reason.
Figure 25 shows the decay velocity with the change of airflow. As it is shown in the
figure, decay velocity increase with the increase of airflow rate.
36 |
Chalmers University of Technology | 50
45
40
35
30
m
c/thgieh
htorf25
3
7
1
26
2
4
8g
g
4
8t
t
g
go
o
t to
on
n
n
n, ,1
1
,
,(
(
1
11
1
(
(/
/
1
24
2
C
CC
C
M
MM
M
C
CC
C
)
))
)
20
15
10
5
0
0 200 400 600 800 1000 1200 1400 1600 1800
time/sec
Figure 27: Froth behavior of different collector concentration at 1 l/min airflow and 900 rpm
stirring rate
The figure 27 shows the froth behavior under the change of collector concentration
at 900 rpm stirring rate, and 1 l/min airflow. With the increase of the collector
concentration, the maximum height increases. However, the growth velocity is not
influenced too much. This is explained by the fact that the airflow is constant, the
bubble amount is constant consequently. Nevertheless, the higher the collector
concentration is, the more particles surface can be changed into hydrophobic, and in
turn the more particles will attach to the air bubbles and can build up the froth to a
higher level. Also table 5 shows that higher the concentration of collector in the
system, longer the single bubble life time will be, as shown in the table below, the
FRT is longer, with the increase of collector concentration that gives higher maximum
froth height.
Table 6 shows a quantitative comparison of the results of different collector
concentration in 900 rpm and 1 l/min airflow.
38 |
Chalmers University of Technology | Figure 33: A summary of the basic model statistics in four parameters; 1 is perfect = 100%.
Figure 33 shows the sumary of the basic model statistics in four parameters as shown
below:
R2 shows the model fit; 0.5 is a model with rather low significance.
Q2 shows an estimate of the future prediction precision. Q2 should be greater than
0.1 for a significant model and greater than 0.5 for a good model.
Model validity is a test of diverse model problems. A value less than 0.25 indicates
statistically significant model problems, such as the presence of outliers, an
incorrect model, or a transformation problem.
Reproducibility is the variation of the replicates compared to overall variability. A
value greater than 0.5 is wanted.
Correct model tuning like removing non significant model parameters or selecting
the appropriate transformation results in higher summary statistics. The best and
most sensitive indicator is Q2.
Note: Model validity might be low in very good models (Q2 > 0.9) due to high
sensitivity in the test or extremely good replicates.
43 |
Chalmers University of Technology | 6. Conclusion
Flotation process is a huge and complex area of research. The chemistry behind this
process is extremely complicated. Froth as discussed earlier play a tremendous role
in flotation. A lot of research has been conducted to study its effect on the flotation
results, especially the froth stability attracts an abundant amount of attention from
the researchers.
This work shows that stirring rate, airflow and collector concentration all can
influence the froth stability.
The repeatability tests of some of the parameters show good repeatability of the
growth velocity, moderate repeatability of the maximum height, but repeatability of
the decay velocity is below the average.
For stirring rate, the froth stability, both the maximum height and growth velocity,
depends on the airflow. At low airflow, the stirring rate is not the dominant
parameter. The froth stability doesn’t change along with the stirring rate. That is
because in this case, the amount of air bubbles is of importance instead of surface
area of the bubbles and also the amount of the hydrophobic particles is constant, the
amount of particle stabilized bubbles are the same in different stirring rates. Also the
air recovery is constant, means the amount of unburst bubble in the column is
constant, which is another explanation of this fact. When the airflow exceeds the
critical value, which is 2l/min predicted by MODDE, the stirring rate starts to affect
the froth stability. The higher the stirring rate is, the higher the maximum height and
the faster the growth velocity is. This is because in this case the surface area of the
bubbles play crucial role, and the higher the stirring rate is, the smaller the bubbles
are, and consequently larger the surface area of the bubbles in the cell. That can
attract more hydrophobic particles, therefore the froth is more stable.
Airflow can affect both the maximum height and the growth velocity. They both
increase, when the airflow is increased. That is because the amount of the bubbles
increase as the airflow is increased, and more hydrophobic particles can load on to
the surface of the bubbles. However, when the airflow comes to a certain number,
the maximum height and the growth velocity will not change with further increase of
the airflow and achieve peak values, respectively. That is because the constant
concentration of the collector in different airflow changing lead to the constant
amount of hydrophobic particles in the cell. There would not be enough particles to
stabilize the excessive bubbles, which would coalesce and collapse in the system.
The collector concentration can significantly influence the froth height, but not so
much on the growth velocity. The froth height can increase along with the increment
of the collector concentration. Nevertheless, the grow rate remains at the same level
with the change of the collector concentration. This is explained by the fact that the
47 |
Chalmers University of Technology | airflow is constant, the bubble amount is constant consequently. Nevertheless, the
higher the collector concentration is, the more particles surface can be changed into
hydrophobic, and in turn the more particles will attach to the air bubbles and can
build up the froth to a higher level.
The experimental design by MODDE software shows that the experimental results fit
the model very well and the prediction out of MODDE is trust worthy. Some of the
prediction of the MODDE can verify the results of the experiments.
The prediction of the MODDE shows both the maximum height and growth velocity
come to the peak value when the airflow and the stirring rate are the highest out of
the experimental parameter set up. But for air recovery the peak value is in the
region where the highest stirring rate and the lowest airflow present. This is because
the hydrophobic particles content wouldn’t change with the increasing of the
amount of the bubble when the airflow is increased. The excessive bubbles would
not be stabilized by the particles and collapse, consequently there would be smaller
amount of unburst bubbles in the system. The higher the stirring rate is, the larger
the surface area of the bubbles would be, and more hydrophobic particles would
attach to the surface, and preventing the bubbles to collapse. MODDE also give a
well prediction of the decay velocity.
48 |
Chalmers University of Technology | Abstract
Inter Terminals AB store oil products for their costumers and need to store them at a certain
temperature with the help of heat exchangers depending on the product. The current heat
exchangers at Inter Terminals are old and does not function efficiently in terms of energy and
costs. The heat exchangers are used to heat oil products in three caverns and a group of four
tanks. The idea is to have heat exchangers for the caverns and for the group of tanks to keep
the oil products at the desired temperature. This assessment evaluates and compares new heat
exchangers for Inter Terminals, which is located in Port of Gothenburg. The companies
responsible for the proposal of the new heat exchangers were Alfa Laval AB, ViFlow AB and
GB Tank AB. Two of the heat exchangers were made of carbon steel and one of stainless
steel. The comparison and evaluation is done through the methods life cycle assessment
(LCA) and life cycle cost (LCC) and three different heat exchangers were compared and
evaluated in terms of environmental and economic aspects. The purpose of this evaluation and
comparison is to help Inter Terminals decision-making to invest in new heat exchangers
which they are in need of. The life cycle assessment was conducted by using OpenLCA,
which is a software tool for calculating environmental impacts. The environmental impacts
categories considered in the assessment were global warming potential, human toxicity,
depletion of abiotic resources and acidification. The weight and the material which a heat
exchanger is made of were two major factors contributing to the environmental impacts.
The results of the LCC were analyzed to calculate the payback time for each heat exchanger.
This was done by comparing the operating costs for the current heat exchangers and the new
ones. This indicated the best alternative of the new heat exchangers for Inter Terminals from
an economic point of view.
Sensitivity analyses were also conducted regarding recycling rates, changes in energy prices
and the percentage amendments of the energy sources. This lead to changes related to the
environmental impacts and the payback time.
The conclusion of this is assessment is that heat exchanger 2, purposed by ViFlow, is best
suited for Inter Terminals and is also manufactured by ViFlow in Örnsköldsvik. It is a shell
and tube heat exchanger made of carbon steel and has the least environmental impacts and has
the shortest payback time compared to the other alternatives.
Keywords: Heat exchanger, Life cycle assessment (LCA), Life cycle cost (LCC), Stainless
steel, Carbon steel, Environmental impact. |
Chalmers University of Technology | 1. Introduction
Inter Terminals is one of the largest independent storage providers of liquid oil products in
northern Europe, and the largest in Scandinavia, with more than 4.25 million cubic meters of
storage capacity (approximately 27 million barrels) located across 16 terminals. Inter
Terminals has four terminals in Sweden along both the east and west coasts, which are
important trading routes for petroleum products. The function of the facilities is as strategic
storage and mixing for the transshipment of petroleum products. They are also an important
part for continental distribution of both petroleum and petrochemical products (Inter
Terminals, 2016).
The Gothenburg terminal is located in the port of Gothenburg, Scandinavia´s largest port and
has easy access to North Sea, - and the Baltic region. It also has convenient access to rail and
road networks and is directly linked via pipelines to three refineries. At Inter Terminals in
Gothenburg products such as fuel oil, heavy oil, diesel, jet fuel and gasoline among many
other are stored and handled. The oil products are stored in both tanks and caverns depending
on the type of the product. These tanks and caverns are using heat exchangers to adjust the
temperatures for the different types of oil products to the requested level (Inter Terminals,
2016). A heat exchanger is a device used to transfer heat between one or more fluids. The
fluids in the heat exchanger can be single or two phase and are often separated by a solid wall
to prevent mixing. This depends on what type of heat exchanger is used, so the fluids may in
some cases be in direct contact (Brogan, 2011).
Inter Terminals AB Sweden stores oil product for their customers in order for them to make a
business in the oil business. The customers usually buy a large amount of oil for a cheap price
and stores the oil at Inter Terminals to sell it later when the demand and price is higher. Since
more countries around the world strive for a carbon free society, Inter Terminals are
interested in how much emissions their equipment is emitting from a lifecycle perspective.
Societies in general wants to make a profit in the oil market, since it still one of the major
markets thus the reason for storing oil.
This study concerns heating up oil products in three caverns and four tanks at the Inter
Terminals facility in Gothenburg. The three caverns could each contain approximately 50 000
m3 of oil products and each of the tanks could hold up to approximately 10 000 m3. The oil
products are arriving at the port of Gothenburg with ships and are then pumped to caverns or
tanks at Inter Terminals. The oil products are stored at Inter Terminals until the customers
want it delivered. The customers also specify which properties the oil products should have.
The temperature of the oil products is important when it is time to deliver it to the customers.
To obtain the desired outlet temperature of the oil Inter Terminals use shell and tube heat
exchangers. The current heat exchangers, which date back to the 70-80s, are not up to date
and need to be replaced in order to increase the heat transfer between the fluids, thus
becoming more energy efficient. The current heat exchangers that Inter Terminals uses to heat
up the oil products are so called water to water heat exchanger and are not constructed to heat
heavy oil. The purpose of the heat exchanger was from the beginning to heat the water bed,
which subsequently warmed the oil. Using the wrong medium impairs the heat exchangers
capability through coatings and fouling on the tubes, along with the high age of the heat
exchangers makes the heat transfer from the water to the oil poor. The current heat
exchangers cannot handle the products that Inter Terminals is storing at the moment to reach
the desired temperature in an effective way without high costs due to long running time. The
current heat exchangers are running at a high cost, due to the poor heat transfer from the
warm water to the oil products in the caverns. For Inter Terminals this is a problem regarding
3 |
Chalmers University of Technology | both economic and the environmental impacts due to inefficient use of energy in the process.
This leads to a bigger carbon footprint than necessary.
In order to explore what type of heat exchangers are suitable to replace the current ones, life
cycle assessment (LCA) and life cycle cost (LCC) have been conducted. These methods will
give detailed information about the environmental and economic impacts and helps Inter
Terminals in the decision-making to determine which alternatives is best suited for their
process. The geographical boundaries are set for Sweden in this study and the time coverage
is set for 30 years for the heat exchangers and 100 years for some of the impact categories.
The aim of this study thus is to review different types of heat exchanger alternatives for Inter
Terminals Sweden AB, regarding economic and environmental aspects. The study will serve
as a basis for Inter Terminals in terms of decision-making for future investments of new heat
exchangers. Therefore, this master thesis will answer these following research questions:
Using LCA and LCC, which heat exchanger is best suited for Inter Terminals
concerning the environmental and economic aspects?
Is there a trade-off between the environmental and economic aspects?
2. Overview of the Technology
Unfortunately, no previous LCA or LCC on heat exchangers were found. However, there
were several LCAs on stainless steel regarding the production and the emissions. According
to International Stainless Steel Forum (ISSF), the CO emission is 2,9 ton CO /ton stainless
2 2
steel. This value will be different depending on the chosen electricity mix since it has a major
impact. The previous LCA studies about steel production were mainly used in order to
understand and model the steel production in this study. Values concerning the energy need in
different steps of the steel production was considered, but in this thesis the Swedish electricity
mix was used instead.
The heat exchangers that are used for the caverns are stationed above ground and each cavern
has its own. The oil in the cavern is pumped up above ground and into the heat exchanger
where warm water transfers its thermal energy to the colder oil and heats it. The oil is then
pumped down into the cavern again and creates stirring and mixing of the oil inside the
cavern, which slightly increases the heat transfer to the surrounding oil. The tanks instead use
pipes in the bottom of the tank with warm water flowing inside them to warm the oil.
The caverns at Inter Terminals are unclad, which means that the oil is stored directly to the
mountain. This kind of storage builds on the principle that oil is lighter than water and that
they do not mix with each other. The caverns are located so that the highest product level is
located at least five meters below the natural ground water. The pressure of the groundwater
in the surrounding mountain will allow the oil to stay in the cavern and prevents it from
leaking throughout cracks in the mountain. Groundwater continuously flows into the cavern
through the cracks and forms a water bed that the oil floats on. The water bed is maintained
constant just above the bottom of the cavern (Avveckling av oljelager i oinklädda bergrum,
2003). Inter Terminals has its own internal water piping system and they are using municipal
drinking water in their pipes. In order to heat Inter Terminals water system, they are buying
hot water from Göteborg Energi and use a heat exchanger to transfer the thermal energy from
the warmer water to the colder water. When the water within Inter Terminals system has been
4 |
Chalmers University of Technology | heated it is transported through pipes to the heat exchangers at the caverns or to the tubes at
the bottom of the tanks to heat the oil products to the desired temperature.
The assumptions throughout the study related to the LCA and LCC are described in the
inventory analysis. Assumptions has been made for each process for the LCA and LCC and
have been explained.
2.1 Heat Exchanger
A heat exchanger is a device used to transfer heat between one or more fluids. The fluids in
the heat exchanger can be single or two phase and are often separated by a solid wall to
prevent mixing, but it´s depending on what type of heat exchanger that is used, so the fluids
could in some cases be in direct contact. When discussing heat exchangers, it is important to
provide some form of categorization. This is mainly done through two approaches. First, to
consider the flow configuration within the heat exchanger, and the second approach is based
on the classification of equipment type by construction. There are four types of basic flow
configurations; counter flow, concurrent flow, crossflow and hybrids. These types of flow
configurations depend on the purpose of the heat exchanger. The classification of the heat
exchangers is mainly classified by their constructions (Brogan, 2011).
The heat exchangers that are interesting and will be focused on in this assessment are shell
and tube heat exchangers (tubular) and plate heat exchangers. They are the most common
types of heat exchangers in the industry (Brogan, 2011). The manufacturers contacted in this
study have designed the proposed heat exchangers based on the specifications from Inter
Terminals and their own knowledge. The type of flow and characterization of the heat
exchangers have been designed by the manufacturers individually.
2.1.1 Shell and Tube Heat Exchanger
Shell and tube heat exchanger belongs to the category of heat exchangers called tubular and
are popular because of the wide range of pressure and temperature. It consists of a number of
tubes mounted inside a cylindrical shell. It functions in the way that two fluids can exchange
heat, one fluid flow through the tubes while the other fluid flows outside the tubes. This
makes the fluids to exchange heat. There are four major parts in a shell and tube heat
exchanger; front end, rear end, tube bundle and shell (Brogan, 2011).
Figure 1. Shell and tube heat exchanger (Brogan, 2011)
A typical shell and tube heat exchanger have the design as in figure 1, with the four major
parts included. The front end is where the fluid enters the tube side of the exchanger, the rear
5 |
Chalmers University of Technology | is where the subside fluid leaves the exchanger or is returned to the front header. The tube
bundle comprises the tubes, tube sheets, baffles and tie rods etc. to hold the bundle together.
Lastly, the tube bundle is contained in the shell. Compared to the plate heat exchanger shell
and tube are often less expensive to buy and could be used in systems with higher operating
temperatures and pressures. Another advantage with tubular heat exchangers is that leaking
tubes are easy to locate and plug (Brogan, 2011).
2.1.2 Plate Heat Exchanger
Another popular type of heat exchanger is the plate heat exchanger, which separate the fluids
exchanging heat by plates. It is often used in cryogenic and food processing industries, but is
also being used in the chemical industries due to its ability to handle more than two streams.
This type of heat exchanger consists of two rectangular end members which holds together
several pressed rectangular plates with holes in the corners for the fluid to pass through
between the plates (figure 2). Plate heat exchangers´ major advantage compared to shell and
tube is that they are easier to take apart for cleaning and maintenance. The heat transfer
efficiency is also higher for a plate heat exchanger and demands less working space (Brogan,
2011).
Figure 2. Plate heat exchanger (Brogan, 2011)
2.1.3 Carbon Steel and Stainless Steel
Heat exchangers, whether they are shell and tube or plate, are mainly constructed with
variations of stainless steel or carbon steel. Steel is an alloy made of carbon and iron, and
depending on the grade of the steel the amount of carbon is varied. Although carbon is the
key alloy element for iron other materials are important to add to obtain various desirable
properties of the steel. For carbon steel, where carbon as said is the main alloying element, the
properties are mainly defined by the amount of carbon that usually is between 0.2% and 2.1%
(pearlitesteel, 2015). A higher amount of carbon increases the strength and hardness of the
steel. For stainless steel chromium and nickel are important alloying materials, the chromium
is used to prevent corrosion and staining. The nickel is added to increase toughness and
combined with a reduced amount of carbon the nickel improves the weldability of the steel
(Outokumpu, 2016).
The main difference between stainless and carbon steel is that the carbon steel can corrode
while stainless steel is protected against corrosion. However, carbon steel has higher thermal
conductivity then stainless steel (pearlitesteel, 2015).
6 |
Chalmers University of Technology | 3. Method
3.1 Life Cycle Assessment (LCA)
Life cycle assessment (LCA) is a method to analyze the environmental impacts associated
with a product or service throughout its life from cradle to grave. The method studies the
energy and material flows for the product and system from raw material extraction, through
production and use, to disposal (Baumann and Tillman, n.d.) As seen in figure 3, life cycle
assessment is carried out in four different steps, goal and scope definition, inventory analysis,
impact assessment and interpretation (UNEP, 2015).
Figure 3. LCA outline with the four steps that is required for a LCA (UNEP, 2015)
A LCA starts with the definition of the goal and scope of the study, which establishes the
context of the study and answers the question how and to whom the results are communicated
to. This LCA step should include specific information such as functional unit, system
boundaries, impact categories, assumptions and limitations, which serve as guidance for the
subsequent analysis.
The goal and scope definition is followed by the inventory analysis which contains all the
relevant data for the LCA study. An inventory of flows, both from and to the nature for the
product system, within the system boundary is created in this step. The inventory includes
inputs of water, energy and raw materials and releases of emissions to air, land and water.
Inventory analysis is followed by the impact assessment; which main purpose is to evaluate
the potential environmental impacts based on the results of the inventory analysis. Significant
impact categories for the product system are selected and all the inventory data are assigned
to the specific impact category that it addresses.
The final step in the LCA is the interpretation, where the results are presented in a
comprehensive way and evaluated regarding the opportunities to reduce material use, energy
and environmental impacts of the product or service that is being studied. The interpretation
looks at each step of the products life cycle in order to evaluate, identify and quantify the
information from the result (Baumann and Tillman, n.d.).
3.1.1 OpenLCA
OpenLCA is a software that is used for conducting Life Cycle Assessments (LCA) and
Sustainability Assessments. The software is developed by GreenDelta, an independent
sustainability consulting and software company (Ciroth, 2016). OpenLCA is used to simplify
all the calculations that must be made for the life cycle assessment.
7 |
Chalmers University of Technology | 3.2 Goal and Scope
The goal of this study is to conduct a LCA and a LCC to serve as a basis for Inter Terminals
Sweden AB in Gothenburg, regarding their decision-making replacing their old heat
exchangers with new ones. Their current heat exchangers are from the 70-80s and are not
efficient enough at the moment in terms of energy efficiency, costs and thermal capacity.
Therefore, this study will analyze the current heat exchangers and compare them to the
alternatives on the market regarding environmental and economic aspects. The LCA will
cover the environmental aspects, while the LCC focus on the economic aspects.
The goals of the two analyses are to:
Analyze the environmental impact in terms of carbon footprint and other
environmental impacts of a heat exchangers lifecycle from cradle-to-grave.
Conduct a life cycle cost analysis on the proposed heat exchanger alternatives for Inter
Terminals Sweden AB.
Evaluate if there is a trade-off between the environmental and economic aspects for
the heat exchangers.
The target audience of this study is mainly Inter Terminals Sweden AB, but it also includes
the steel industry and the heat exchanger industry.
The results of this study are intended to assist Inter Terminals Sweden AB decision-making in
which heat exchangers to invest in. The results can also be used for the steel industry to locate
hot spots in the process concerning energy consumption and environmental impacts.
3.2.1 Functional Unit
The functional unit is used in the LCA study to describe the function of the system under
study. The functional unit that will be used in this study is the amount of energy needed to
keep the oil products at the desired temperature over the life time of the heat exchanger.
3.2.2 System Description
Figure 4 illustrates the processes that are included in the study, the material flow from cradle
to grave. The boxes represent different processes and the arrows are illustrating the direction
of material flow within the system. Landfill is not included in the calculation. The energy
mixes included is first of all the Swedish electricity mix, but if that is not available in
OpenLCA global and rest of the world is used. The chemicals, gases and other substances
used during the life cycle of the heat exchangers are not included due to lack of data.
8 |
Chalmers University of Technology | Figure 4. LCA flowchart for a heat exchanger
3.2.2.1 Extraction
The first step in a LCA is the extraction phase, which includes all the flows related to the
extraction of the raw materials. This step in this assessment refers to the materials and energy
needed to produce stainless steel and carbon steel which will be used in the manufacturing
process for the heat exchangers later on in the system.
In the extraction phase, all the materials needed are extracted. This phase will be different
depending on the material of the heat exchanger, if it’s made of stainless steel or carbon steel.
The transports within the extractions site are included in the inputs related to the extraction
processes in OpenLCA.
3.2.2.2 Steel Production
Stainless steel and carbon steel are produced in the same way; the only difference is the
material composition. Figure 5 below illustrates one route of processes and flows regarding
steel production, this route is considered in this study and is the most common one.
9 |
Chalmers University of Technology | Sinter Plant
The sintering process is a pre-treatment step for iron making where sinter is produced for the
blast furnace step (Ispatguru, 2016). Before the actual sintering process different ferrous
containing materials are mixed. These substances can be derived from various ores mainly
iron (Sinter plants, 2001). This mix is blended with coke particles, limestone and heated in
order to produce a semi-molten mass. This molten mass is solidified to porous sinter pieces
that are used in the blast furnace process (Ispatguru, 2016) (Industrial Efficiency Technology
& Measures, 2016).
Blast Furnace
Sinter and coke are introduced to the blast furnace where hot air is blasted from the bottom of
the furnace. Carbon monoxide is then formed when oxygen in the air is combusted with the
coke. The carbon monoxide flows up through the blast furnace and removes the oxygen in the
sinter leaving iron. The high temperature in the furnace melts the iron and the liquid iron is
tapped into ladles where the iron liquid is transported to the steel furnace also called basic
oxygen steelmaking process (EEF, 2016).
Basic Oxygen Steelmaking
The liquid iron from the blast furnace is the main raw material in the basic oxygen
steelmaking process, the rest is balanced with steel scrap. The liquid iron and the steel scrap is
inserted into the basic oxygen furnace, where nearly pure oxygen at high temperature is
blown in (Steel, 2016). Through oxidation unwanted elements are separated from the metal
and left is molten steel that is tapped into ladles. Alloys could be added to the molten steel
before its goes on to the casting machine where the steel is solidified (EEF, 2016).
Continues Casting
Molten steel from basic oxygen furnace is inserted in the continuous casting machine where
the steel passes through a series of rolls and water sprays. This process ensure that steel is
rolled into right shape and at the same time fully solidified. The steel is straightened and cut
into the required length at the end of the process, forming so called slabs, billets and blooms.
These are transported to the hot rolling mill in order to produce steel products (EEF, 2016).
Hot Rolling Mill
The slabs, billets and blooms from the steelmaking process are transported to the hot rolling
mill. These semi-finished products can be classified into flat or long products based on their
shape. Slabs are used to roll flat products, that are used to manufacture steel components for
heat exchangers. The steps that are followed in the hot rolling mill are the use of different
finishing steps depending of the desirable shape and dimension of the finished product.
Examples of finished products could be plates, large and small tubes that are used in the heat
exchanger manufacturing (EEF, 2016).
3.2.2.3 Heat Exchanger Production
The heat exchangers are manufactured in this process, by using either stainless steel or carbon
steel and adding energy and transportation. The materials that the heat exchangers are made
of depends on the type of heat exchanger and the company manufacturing it. In this
assessment only two materials are considered, stainless steel and carbon steel.
11 |
Chalmers University of Technology | 3.2.2.4 Use Phase
In the use phase, the products are being used. Some products require energy during their use
phase and that needs to be accounted for. Heat exchangers do not require any energy in form
of electricity, but the heated water that runs through the heat exchanger is heated by various
energy sources. The use phase for this assessment will be the energy that is needed to heat up
the water that runs through the heat exchangers in order to keep the oil products at the desired
temperature.
3.2.2.5 Recycling
Steel is the most recycled material, even more than all other materials combined, and its
properties makes it possible to continually recycle the steel without any degradation in
performance and quality (Steel, 2016). Collected metal scrap is transported to recycling
facilities where it is checked, sorted and processed to meet the steel producer’s quality
requirements (Stena Recycling, 2016). Recycled metal scrap is then transported back to the
steel production facility were its reintroduced in the steel production.
3.2.3 Time Coverage
The study will take into account a lifetime of 30 years for the heat exchangers, both for the
LCA and the LCC. The impact assessment will cover a time of 100 years for the impact
categories global warming potential and human toxicity.
3.2.4 Geographic Coverage
The geographical coverage of the assessments is set for Sweden, both for the LCA and LCC.
The origin of the oil products heated in the heat exchanger are not taken into account, thus
neglected in the study.
3.2.5 Impact Categories
The impact categories that this study includes and are global warming potential, human
toxicity potential, acidification potential and depletion of abiotic resources. Inter Terminals is
interested in the carbon footprint for the heat exchanger, therefore global warming potential is
chosen as an impact category. Global warming potential considers the release of so called
greenhouse gases to the atmosphere and is weighted against the effect same amount of carbon
dioxide has, thus the unit kg CO -equivalent (Green Guide to Specification BRE Materials
2
Industry Briefing Note 3a: Characterisation, 2005).
Depletion of abiotic resources is chosen as an impact category because heat exchangers
mainly contains different metals. Depletion of abiotic resources focus on the extraction of
scare minerals and metals and the depletion factor is determined by the extraction rate and the
remaining reserves. The depletion factor for the studied mineral/metal is compared to the
factor for Antimony (Sb), which is used as a reference case. The unit for abiotic resources is
hence kg Sb-equivalent (Green Guide to Specification BRE Materials Industry Briefing Note
3a: Characterisation, 2005).
Human toxicity is chosen to evaluate how emissions from substances used throughout the
lifecycle impacts human health. The human toxicity potential depends on the fate, exposure
and effects of the toxic substance for an infinite time horizon and is expressed by the
reference unit, kg 1,4-dichlorobenzene (DCB) -equivalent (Green Guide to Specification BRE
Materials Industry Briefing Note 3a: Characterisation, 2005). The reason to include this
impact category is to indicate that some products are harmful to humans even though the
12 |
Chalmers University of Technology | product itself has no danger to humans directly during its use phase. However, the
manufacturing of the products could include processes which are harmful to humans.
Throughout a lifecycle for a heat exchanger a lot of energy and substances are used, which in
different ways contribute to acidification. Processes such as road transportation and the
energy used are examples of processes that contributes to acidification (SO , NO and NH ).
x x 3
Therefore, is this impact category analyzed in this assessment. This impact category uses kg
SO -equivalent as reference unit (Green Guide to Specification BRE Materials Industry
2
Briefing Note 3a: Characterisation, 2005).
3.3 Life Cycle Cost (LCC)
Life cycle costing (LCC) is a tool that concerns all the costs that arise from an investing
decision and can be used to evaluate complete processes, systems and objects throughout its
life cycle. LCC is used to assist and improve the decision-making process. LCC is defined as
the present value of the total cost of an asset over its operating life, including initial capital
cost, installation costs, operating costs, maintenance costs and the cost or benefit of the
eventual disposal at the end of its life. That is the total cost that the project will impose
throughout the whole of its life (Constanza et.al P 36-37, 2013).
In this assessment LCC will be used to calculate the total costs related to the heat exchangers
that are being studied and compared. Since Inter Terminals are also interested in the costs,
LCC seemed like a good method to use for this assessment. Inter Terminals wants to become
more environmentally friendly, but at the same time increase the economic effectiveness. The
LCC will take into account all the costs related to each heat exchanger and compare them to
see which alternative is best regarding the economic aspects. Together with the LCA, it will
be a guideline for Inter Terminals decision making in the future for the investment of heat
exchangers.
The study will include the capital cost, installation cost, maintenance, operating cost and
recycling profit of the three heat exchangers. The operating cost for the current heat
exchangers at Inter Terminals will also be included to enable comparison of possible
operation cost savings if investing in new heat exchangers. This LCC is conducted as a
cradle-to-grave. The time coverage for this LCC is 30 years, its estimated in consultation with
the three companies and it´s said that the heat exchangers would function well at least for 30
years. In this study Inter Terminals are interested in four heat exchangers, three for their
caverns and one for their group of four tanks. This LCC will just consider costs regarding
three heat exchangers, thus the heat exchanger for the tanks will be excluded. The heat
exchangers proposed by the companies are designed for the caverns volume of approximately
50 000 m3. The companies designing the heat exchanger did not consider that the volume of
the tanks approximately is a fifth of the caverns size. This would have an impact on the size
of the heat exchanger for the tanks, thus direct impact on all of the costs. The tanks at Inter
Terminals are not using a heat exchanger right now, as said they are using tubes, with hot
water, in the bottom of the tanks to heat the oil. The heat exchanger planned for the tanks is
intended to only be used to boost the heating in cases of big temperature drops or used to heat
the oil in order to simplify emptying of the tanks. Because they do not use a heat exchanger
for the tanks today it is difficult to compare the energy used today and the energy
consumption after a heat exchanger is installed. It becomes difficult to estimate how many
hours per year this heat exchanger would be used, and it is not correctly dimensioned. This
would not give a good approximated value of the energy consumption, so it would not make a
13 |
Chalmers University of Technology | relevant analysis of the costs for this exchanger and thus not relevant to the LCC analysis.
Because of that the fourth heat exchanger is excluded in this cost analysis.
3.4 Sensitivity Analysis
The sensitivity analysis method has been applied to determine the difference in environmental
impact using different recycling rates (60, 80 and 100 %) of steel scrap to decrease the
extraction and use of virgin metals.
The time coverage for the life cycle assessment is 30 years for the heat exchangers, meaning
that the use phase is most likely the primary contributor regarding the environmental impact.
Therefore, a sensitivity analysis is conducted on the energy sources for the use phase. This is
performed by increasing the amount of energy, in percentage (20 and 40%), for each energy
sources separately, which decreases the other energy sources. This was done in order to
evaluate how the environmental impacts were affected by the energy sources during the use
phase.
A sensitivity analysis was conducted for the life cycle cost regarding changes of the energy
price. This will affect the payback time for the analyzed heat exchangers. This is an essential
analyze, because the energy price would most likely to change over the next 30 years which is
the life cycle of the heat exchangers. This sensitivity analysis is conducted by using the
calculated energy price for 2016 as a starting point and then alter the energy price by
increasing and decreasing it with 10, 25 and 50%.
3.5 Data Sources
Ecoinvent is one of the most used LCA-databases worldwide and is regularly updated.
Ecoinvent is a transparent and consistent life cycle inventory database that provides process
data for thousands of products, in order to simplify decision-making based on environmental
impact (GmbH, 2016). The inventory data are derived from international industrial life cycles
and contains for instance data regarding energy supply, resource extraction, material supply,
chemicals, metals, waste management services and transports services (Openlca, 2016).
Except Ecoinvent, other data sources were used such as Chalmers database and the companies
involved in the assessment (Inter Terminals, Göteborg Energi, Stena Recycling, Alfa Laval
AB, ViFlow AB and GB Tank AB).
3.6 Delimitations
Every flow related to the whole process of the LCA is based on Swedish values such as
energy and extraction locations, but if it’s not available in ecoinvent then global and rest of
the world flows are used for the calculations. Also, the technical details about the heat
exchangers are not included in this process, only the materials they are made of. Due to lack
of data the chemicals and other substances used in the production of the heat exchangers are
not included in this assessment.
14 |
Chalmers University of Technology | 4. Inventory Analysis
Three companies (Alfa Lava AB, GB Tank AB and ViFlow AB) were contacted and willing
to help with the thermal design and dimensioning of the heat exchangers, in order to suggest
one to fit Inter Terminals specifications and facility. In table 1 below the specifications from
Inter Terminals are stated, which requirements the proposed heat exchangers had to manage.
Table 1. Technical requirements from Inter Terminals.
Inlet temp. water 90°C
Outlet temp. water >76°C
Inlet temp. oil 46-65°C
Flow (water) 30 m3/h
Flow (oil) 350 m3/h
Type of oil ISO VG equivalent
Density (oil) 0,9-0,97 kg/m3
Viscosity (oil) at 50°C 800 Cst
Specific heat capacity ~1,967 KJ/kg*°C
(oil)
Volume (caverns) 48 500 -59 355 m3
Volume (tanks) 9 000 – 11 800 m3
Dimensional limits >10m
The heat exchangers are expected to be designed so they manage to heat a volume of 50 000
m3 oil 1 degree Celsius per day. It is not required that the total volume of 50 000 m3 oil
should pass through the heat exchanger in one day. The oil that is passed through the heat
exchanger is heated more than one degree Celsius, then pumped back down into the cavern
heating the surrounding oil which leads to an overall heating of one degree Celsius for 50 000
m3.
This chapter gives an explanation to all processes included in the life cycle assessment and
the three heat exchangers that were obtained are analyzed, compared and presented in this
chapter. They have been divided and named heat exchanger 1, 2 & 3.
4.1 Heat Exchanger 1
The proposed heat exchanger that Alfa Laval Lund AB provided is a plate heat exchanger of
the model T35-PFM and will be referred to as heat exchanger 1 in this assessment. It is
mainly made of stainless steel; therefore, the assumption has been made that the whole heat
exchanger consists of stainless steel. Other parts of the heat exchanger, like gaskets, that are
not made of stainless steel accounts for a very small mass of the heat exchanger. Table 2,
gives a short summary the heat exchanger.
Table 2. Heat exchanger 1 (Alfa Laval).
Name: Heat Exchanger 1
Material: Stainless Steel (304)
Weight: 8810 kg
15 |
Chalmers University of Technology | 4.1.1 Extraction
The first process in the life cycle assessment is the extraction. The extraction flows for heat
exchanger 1 are from Rest of the World (RoW) and Global (GLO) due to lack of data for
Sweden. Stainless steel is mainly made of iron with small percentages of other elements such
as coal, manganese, chromium, nickel, ferrosilicon, phosphor, sulfur and the rest is balanced
with iron (Table 3). There are many different types of alloys for stainless steel, the one used
in this heat exchanger is called alloy 304. The extraction of the materials is assumed to occur
in the Swedish city Kiruna in northern Sweden. The amount of energy required and the
transports within the extraction site are included in the flows in OpenLCA. Sulfur and
phosphor are not included in the calculations in OpenLCA due to lack of data and the small
amounts, therefore they have been excluded.
Table 3. The composition of stainless steel (304) in percentages (Rjsales, 2016).
4.1.2 Stainless Steel Production
The second process in the LCA is the stainless steel production. The process to manufacture
stainless steel is complicated and requires multiple steps, but the total amount of energy
needed for these steps are 52,4 MJ/kg stainless steel. This amount is actually for steel in
general and not specifically for stainless steel, since the procedure is the same. Therefore, the
assumption has been made that the amount of energy is the same for stainless steel as it is for
steel. The energy is coming from the Swedish electricity mix which is available in ecoinvent.
The coke making process, which is a part of the process for stainless steel production,
requires 1,6 kg hard coal and 5 MJ to produce 1kg of coke. This amount of energy accounts
for 10% of the energy demand in the blast furnace and basic oxygen furnace steps which is a
part of the stainless steel production (Industrial Efficiency Technology & Measures, 2016).
Beside these steps, there is also a hot rolling process which require 2,4 MJ/kg (Industrial
Efficiency Technology & Measures, 2016). This means that the energy demand for the whole
stainless steel production is 52,4 MJ/kg, while 2,4 MJ/kg is from the hot rolling and 50 MJ/kg
from the other process for stainless steel production. It was assumed that only the biggest
steps in the production was included, which is; coke making, hot rolling, blast furnace and
basic oxygen furnace due to lack of data for the other steps. The iron that is needed to produce
stainless steel is sintered iron (Ispatguru, 2016).
4.1.3 Heat Exchanger Manufacturing
As seen in table 2, the heat exchanger weights 8810 kg, and is assumed to be made of
stainless steel only. The heat exchanger is manufactured by Alfa Laval at their manufacturing
site which is located in Lund. The energy used in their facility is coming from three different
sources; electricity, heat district and natural gas. The natural gas and heat district is from
RoW, since no data for Sweden was available in ecoinvent. The energy amounts were
obtained from Alfa Laval, but it is not precise since they do not have the exact amount of
energy for each heat exchanger manufactured. Therefore, the amounts are generic and have
been obtained by looking at their annual energy production divided by the amount of steel
treated at their manufacturing site. These numbers were the best data that could be achieved,
even if it’s not only stainless steel they are working with since there were no specific data for
16 |
Chalmers University of Technology | each heat exchanger regarding the energy consumptions (Lundgren, 2016). The lifetime of the
heat exchanger is assumed to be 30 years.
4.1.4 Use Phase
Heat exchangers are not electronical devices, meaning that they do not consume any kind of
energy or emit emissions during the use phase. However, Inter Terminals are getting hot
water from Göteborg Energi to heat up the water in their internal water system to heat the oil
products via the heat exchanger. Göteborg Energi uses different kind of energy sources to
heat the water to a certain temperature which Inter Terminals then use. By knowing the flow
of the hot water, the amount of energy needed to maintain the oil products at a certain
temperature can be calculated. The majority of the energy is coming from burning pellets, but
also from natural gas and heating fuel. The data were accessed from Göteborg Energi annual
production for Skarvik, which Inter Terminal is a part of in Port of Gothenburg. Due to lack
of data in ecoinvent, RoW energy mix was used for the calculations in OpenLCA (Lilienberg,
2014).
Göteborg Energi Energy
Göteborg Energi uses three different types of energy sources to heat the water for Inter
Terminals; pellets, natural gas and heating fuel (table 4).
Table 4. Energy sources and the energy amount for Göteborg Energi without losses (Lilienberg, 2014).
Energy Source Amount of Energy Percentage
E01 (heating fuel) 398 MWh = 1432800 MJ 1,5 %
Natural Gas 8190 MWh = 2948400 MJ 31 %
Pellets 17694 MWh = 63698400 67,5 %
MJ
Total 26282 MWh = 94615200 100 %
MJ
Table 4 describes where Göteborg Energi is getting their energy from and the amounts for
each source. The numbers are without losses which occurs later on when the energy is used
and it is a 16% loss. The same percentage will be used for the losses for Inter Terminals
facility, since the transportation way is similar to Göteborg Energis (table 5). This data is only
accountable for Skarvik in Gothenburg Port.
Inter Terminals Energy Demand Calculations
These calculations accounts for the use phase for all the heat exchangers, since the amount of
energy needed is the same.
Göteborg Energi heat up Inter Terminals water from 75 to 90 degree Celsius.
The water flow rate in the heat exchangers are 30 m3/h.
The new heat exchanger has a capacity to heat the oil in the tanks and caverns by 1 degree
Celsius per day, while the old ones could only heat 0,1 degree Celsius per day. Therefore,
the new heat exchangers require only 10% of the time to heat up the oil products
compared to the old ones.
The current heat exchangers are estimated to operate 30% of the year which is 2628
h/year.
17 |
Chalmers University of Technology | The energy loss has been estimated to be 16% from pump, valves, pipes etc, which is the
same value for Götebrg Energi when they produce the energy.
Heating water from 75-90 degree requires 504 kWh/h = 1814,4 MJ/h
This is calculated by a calculator on the website Lenntech (Lenntech, 2016).
Operating time for the new heat exchangers: 10% * 2628 h/year = 262,8 h/year
Amount of energy required to maintain the oil products at the desired temperature without
energy losses: 262,8 h/year * 1814,4 MJ/h = 476 824 MJ/year
Taking the energy losses into account: 476 824 MJ/year * 1,16 = 553 116 MJ/year/heat
exchanger
Table 5. Energy required distributed over different energy sources based on data from Göteborg Energi (data for use of one
new heat exchanger and one year with 16% energy loss) (Lilienberg, 2014).
Energy Source Amount of Energy Percentage
E01 (heating fuel) 8297 MJ 1,5 %
Natural Gas 171466 MJ 31 %
Pellets 373353 MJ 67,5 %
Total 553116 MJ 100 %
For the heat exchanger lifetime (30 years): 553 116 MJ/year * 30 years = 165 934 80 MJ
4.1.5 Recycling
For the recycling, three different recycling rates were calculated in OpenLCA to see how it
affects the result. The recycling rates were 60%, 80% and 100% and it goes back into the steel
production. After the lifetime of the heat exchanger, the materials are collected, sorted and
processed to be used again. These steps require transport and energy. In ecoinvent, there was
a process but it had to be modified to make it suitable for this assessment. All the energy
flows related to the process were added together to see how much energy is needed to collect
and sort one kg of stainless steel. The Swedish energy mix was used for this. The reason for
including a recycling process is that in reality, steel in general has a high recycling rate which
is reused in every equipment. The rest that is not recycled has been assumed to go to landfill.
4.1.6 Transport
The metals used to produce stainless steel are assumed to be extracted in LKAB:s mine in
Kiruna in northern Sweden, because it is one of the largest mining operations in the world.
The extracted metals are assumed to be transported to Outokumpus facility in Avesta.
Outokumpu is a leading company in stainless steel production in Sweden, therefore it is
assumed that the stainless steel for the heat exchanger is produced by them (Jernkontoret.se,
2015). The transport distance from Kiruna to Avesta is 1 160 km.
The stainless steel is assumed to be transported directly from Avesta to Alfa Laval in Lund
where the heat exchangers are manufactured, transports to and from various steel wholesalers
are neglected. The stainless steel is transported 630 km from Avesta to Lund. The heat
exchanger is then transported 262 km from Alfa Laval to Inter Terminals Sweden AB in
Gothenburg.
18 |
Chalmers University of Technology | The heat exchanger is dismantled after its lifetime and is assumed to be transported to Stena
Recycling in Gothenburg, this transportation is neglected because of the short distance. Metal
scrap from the heat exchanger are sorted and processed at Stena Recycling before it is
transported back to the steel production, it is assumed that all the metal scrap goes to
Outokumpus steel production in Avesta where it is recycled. The distance between
Gothenburg and Avesta is 419 km. All transports are assumed to be carried out with EURO 4
lorries (GLO). The transportations can be seen in table 6.
Table 6. Transportation route for heat exchanger 1.
Transport - Heat Exchanger 1
Transport Path (EURO 4 Lorry) Distance (km)
Kiruna-Avesta 1160
Avesta-Lund 630
Lund-Gothenburg 262
Gothenburg-Avesta 419
Total 2471
4.2 Heat Exchanger 2 & 3
The other heat exchangers proposed were from two other companies, ViFlow and GB Tank.
They are both shell and tube heat exchangers and are made of the same material but has
different weights. In this assessment ViFlow´s heat exchanger is referred to as heat exchanger
2 and GB Tank´s heat exchanger 3. Below in table 7 and 8, a brief summary of these heat
exchangers can be seen. Both are made of carbon steel, and its assumed that the whole heat
exchangers are made of it. The weight for heat exchanger 2 is for two heat exchangers, since
ViFlow recommended to use two heat exchangers of the same model in series in order to
maintain the oil products at the desired temperature.
Table 7. Heat exchanger 2 (ViFlow AB).
Name: Heat Exchanger 2
Material: Carbon Steel
Weight: 6400 kg
Table 8. Heat exchanger 3 (GB Tank AB).
Name: Heat Exchanger 3
Material: Carbon Steel
Weight: 22301,4 kg
4.2.1 Extraction
The modelling for these heat exchanger 2 & 3 were exactly the same in OpenLCA as for heat
exchanger 1, except for some different values regarding the extraction. Since these heat
exchangers are made of carbon steel, the composition is different from stainless steel,
meaning that the extraction phase is different. The composition of carbon steel can be seen in
Table 9. As for the stainless steel heat exchanger, phosphorous and sulfur were not included
in the extraction phase in OpenLCA due to lack of data.
19 |
Chalmers University of Technology | Table 9. The composition of carbon steel (Ibrahim, 2010)
C Si Mn Cr Mo Ni P S Cu Fe
0,40 0,31 1,08 0,14 0,005 0,104 0,012 0,043 0,169 Balanced
4.2.2 Carbon Steel Production
The production of carbon steel is the same as for stainless steel production since the
procedure is identical. Therefore, the amount of energy needed to produce carbon steel has
been assumed to be the same as for stainless steel which is 52,4 MJ/kg.
4.2.3 Heat Exchanger Manufacturing
The manufacturing of the carbon steel heat exchangers had different energy values than the
one made of stainless steel. It is still assumed that the whole heat exchangers are made of
carbon steel, but the energy values required are different. For heat exchanger 1 (stainless
steel), Alfa Laval provided data regarding their annual energy consumption and the amount of
material treated which made it easy to calculate and estimate an approximate value of the
energy required to manufacture one heat exchanger. For heat exchanger 2 and 3, no such data
were available, therefore some assumptions were made to calculate the energy needed for the
manufacturing of these heat exchangers. According to ViFlow, the energy needed to
manufacture a heat exchanger is a small percentage of the amount of energy needed for the
carbon steel production. To verify this statement, heat exchanger 1 was analyzed. The energy
for manufacturing the heat exchanger 1 was only 1,4% of the energy needed for the stainless
steel production and since that amount also accounts for the carbon steel, the statement
seemed to be an appropriate assumption (Lindberg, 2016). This value was assumed for GB
Tanks manufacture process as well. To calculate the energy required to manufacture heat
exchanger 2 and 3, the same amount (1,4%) was used. This is also an assumption since it is
not the exact amount of energy needed for the manufacturing of the heat exchanger in reality.
4.2.4 Use Phase
The use phase is exactly the same as for heat exchanger 1, since the energy needed to
maintain the oil products at the desired temperature is the same for all three heat exchangers.
To see the calculations, see chapter 4.1.4.
4.2.5 Recycling
The recycling for heat exchanger 2 and 3 is also the same as for heat exchanger 1.
The metals goes back into the carbon steel production instead of the stainless
steel production. The recycling rates are the same; 60%, 80% and 100%.
4.2.6 Transport
The metals used to produce carbon steel are assumed to be extracted in LKAB:s mine in
Kiruna in northern Sweden. The extracted metals are then assumed to be transported to
SSABs facility in Luleå, where carbon steel is produced. The distance between Kiruna and
Luleå is 343 km. Carbon steel is transported from Luleå to ViFlow in Örnsköldsvik, a
distance of 375 km, where the heat exchanger manufacturing takes place. Transports of
carbon steel to and from various steel wholesalers ViFlow are neglected. The heat exchanger
is transported 873 km from ViFlow to Inter Terminals Sweden AB in Gothenburg.
It is assumed that the heat exchanger after its lifetime are transported to Stena Recycling in
Gothenburg, this transport is neglected because of the short distance. All of the sorted and
20 |
Chalmers University of Technology | processed metal scrap at Stena Recycling are assumed to be recycled and returned to SSABs
steel production in Luleå, a distance of 1 247 km. All transports are assumed to be carried out
with EURO 4 lorries (GLO). The transportations can be seen in table 10.
Table 10. Transportation route for heat exchanger 2.
Transport - Heat Exchanger 2
Transport Path (EURO 4 Lorry) Distance (km)
Kiruna-Luleå 343
Luleå-Örnsköldsvik 375
Örnsköldsvik-Gothenburg 873
Gothenburg-Luleå 1247
Total 2838
The extraction of the materials for heat exchanger 3 also takes place in LKAB:s mine in
Kiruna. The extracted metals are assumed to be transported 343 km to SSAB in Luleå, where
carbon steel is produced. Carbon steel is transported from Luleå to GB Tank AB in Falun, a
distance of 820 km, where the heat exchanger manufacturing takes place. Transports to and
from various steel wholesalers are neglected for heat exchanger 3 as well. The heat exchanger
is transported 461 km from GB Tank AB to Inter Terminals Sweden AB in Gothenburg.
The heat exchangers are assumed to be transported to Stena Recycling in Gothenburg, this
transport is neglected because of the short distance. All of the sorted and processed metal
scrap at Stena Recycling are assumed to be recycled and returned to SSABs steel production
in Luleå, a distance of 1 247 km. All transports are assumed to be carried out with EURO 4
lorries (GLO). The transportations can be seen in table 11.
Table 11. Transportation route for heat exchanger 3.
Transport - Heat Exchanger 3
Transport Path (EURO 4 Lorry) Distance (km)
Kiruna-Luleå 343
Luleå-Falun 820
Falun-Gothenburg 461
Gothenburg-Luleå 1247
Total 2871
4.3 Summary of the Heat Exchangers
A summary of the heat exchangers analyzed in this assessment is presented in the table below
(table 12). It shows the type of material, weight of the heat exchangers and the energy
requirement from each step in the production process. The weight of the heat exchangers
alters for the different alternatives depending on the design made by the manufactures. The
extraction phase was constructed in OpenLCA and includes several different energy flows
which is time consuming to sum up, therefore only the composition of the materials is
included. As mentioned earlier, the composition for stainless steel and carbon steel is
different. For stainless steel the amount of chromium and nickel is higher than for carbon
steel, which can impact the result. The extraction phase is also dependent of the recycling
rates, the higher the recycling rate the less energy is needed from the extraction phase. The
21 |
Chalmers University of Technology | energy needed for all of the heat exchangers is the same regarding the steel production. This
is due to that the procedure is the same for both stainless steel and carbon steel. The different
energy values for the heat exchanger production is due to the data collected. With help from
Alfa Laval the energy required could be estimated. For heat exchanger 2 and 3, assumptions
were made which are explained in detail in section 4.2.3. The energy required for the
recycling process is the same for all of the heat exchangers and was obtained through
ecoinvent. The difference in the transport is due to different locations of companies.
Table 12. Summary of energy flows, material flows and transportation for the analyzed heat exchangers.
Heat Exchanger 1 Heat Exchanger 2 Heat Exchanger 3
(Alfa Laval) (ViFlow) (GB Tank)
Material: Stainless Steel Carbon Steel Carbon Steel
Weight: 8810 kg 6400 kg 22301,4 kg
C (0,035%), Cr C (0,4%), Si (0,31%), C (0,4%), Si (0,31%),
Extraction: (20%), Mn (2%), Mn (1,08), Cr (0,14%), Mn (1,08), Cr (0,14%),
Ni (13%), Si Mo (0,005%),Ni Mo (0,005%),Ni
(0,75%), Fe (0,104%), Cu (0,169%), (0,104%), Cu (0,169%),
(64,22%) Fe (97,79%) Fe (97,79%)
Steel Production: 52,4 MJ/kg 52,4 MJ/kg 52,4 MJ/kg
Heat Exchanger 0,8028 MJ/kg 0,7336 MJ/kg 0,7336 MJ/kg
Production:
Use Phase: 553116 MJ/year 553116 MJ/year 553116 MJ/year
Recycling: 0,144 MJ/kg 0,144 MJ/kg 0,144 MJ/kg
Transport: 2471 km 2838 km 2871 km
4.4 Life Cycle Cost
4.4.1 Capital Cost
The capital costs for the heat exchangers were obtained from the representatively company.
Costs are given for one heat exchanger and then multiplied with the number of heat
exchangers needed. Alfa Laval and GB Tank proposed one heat exchanger per cavern while
ViFlow proposed two heat exchangers in series for each cavern.
4.4.2 Installation Cost
The installation cost for heat exchanger 1 was estimated by ViFlow. For Alfa Laval and GB
Tank no costs could be obtained, therefore the installation costs for heat exchanger 2 & 3
were assumed to be the same as ViFlow´s. This assumption was made in consideration with
relevant staff at Inter Terminals. Costs for building foundations and operation houses for the
heat exchangers are not included in the installation cost, only the actual installation
connecting the pipes are accounted for. The capital and installation costs are a onetime cost
and is not dependent on the heat exchangers lifetime. These costs are illustrated below in table
13, the number of heat exchangers needed is also given. HX in the tables is an abbreviation of
heat exchanger.
22 |
Chalmers University of Technology | PI = Pellets price index bulk (min. 15-ton delivery), annual average calculated from
1
monthly index according to Pelletsförbundet for the year before the current year = 102,0
PI = As above with base year 2013 = 102,0
bas
KPI = Consumption price index according to Statistiska Centralbyrån (Statistiska
1
Centralbyrån, 2016), the average value of the year before the current year = 313,49
KPI = As above with base year 2013 = 314,06
bas
Insertion of values in equation 2 gives an energy price of 907,7 SEK/MWh for 2016.
4.4.5 Effectivity Bonus
The effectivity bonus depends on the return temperature of the water to Göteborg Energi. The
return temperature is on the primary side of the water-water heat exchanger at point B in
figure 8.
Figure 6. How Göteborg Energi heat the water inside Inter Terminals internal water system.
The new heat exchangers are designed to give an outlet temperature of approximately 75°C, if
the inlet temperature of the water is 90 °C and the water flow is 30 m3/h. After passing
through the heat exchanger the water is pumped back to the water-water heat exchanger (point
D, in figure 8), where Inter Terminals internal water system is heated by Göteborg Energis
hot water. The return temperature (point B) has to be calculated in order to obtain the correct
effectivity bonus. The return temperature is weighted against the energy consumption. The
energy consumption and return temperature is measured per hour. For the monthly hours, the
energy consumption is multiplied by the return temperature, the sum of these values is
divided by the month's total energy consumption, giving the energy-weighted return
temperature. The data needed to calculate the return temperature were however impossible to
obtain, since the data for the current heat exchangers are not valid or possible to relate to the
new heat exchangers. In consultation with Inter Terminals the return temperature at the
primary side (point B) was assumed to be the same as the return temperature at the secondary
side (point D), when looking for a whole month.
The return temperature for the current heat exchangers is needed in order to be able to do a
comparison between the operation costs and determine possible cost savings when investing
in new heat exchangers. As mentioned earlier, the current heat exchangers are old and do not
function well, therefore Inter Terminals is currently running the heat exchanger with an
average water flow of 10 m3/h. The reason for this is to let the heated water stay as long as
possible in the heat exchangers and get a higher heat transfer between the water and oil. To be
able to compare the operating costs the return temperature of the water for 10 m3/h must be
converted to a return temperature for 30 m3/h for the current heat exchangers. The inlet and
outlet temperature of the water with a flow rate of 10 m3/h is 94 °C and 75 °C, resulting in a
delta temperature of 19 °C. New delta temperature for a flow rate of 30 m3/h can be calculated
with equation 3 and 4 below.
24 |
Chalmers University of Technology | Table 15. Required energy for new and current heat exchangers with and without energy loss through the system.
Required energy/HX (MWh/yr)
New HX Current HX
With losses (16% pipes, pumps, HX etc.) 153,64 649,33
Without losses 132,45 559,76
The mean operating cost value is calculated for both the new and the current heat exchangers
in order to determine the cost savings regarding operating cost when investing in new heat
exchangers. This is shown in table 16 below.
Table 16. Operating cost for new and current HX and cost savings.
Operating cost
Operating cost (SEK) New HX Current HX Cost saving
Operating cost/year/HX 106 955 548 745 441 791
Total operating cost/year 320 864 1 646 236 1 325 372
Total operating cost 30 years 9 625 927 49 387 095 39 761 168
4.4.5 Maintenance Cost
The maintenance costs include cost of cleaning plates/tubes, changing gaskets and the labor
itself. The cost of traveling and accommodation for the workers is also included in the
maintenance price. It is assumed that the company that manufactures the heat exchangers also
maintains them. The maintenance costs for each heat exchanger are approximate numbers
estimated in consultation with GB Tank, ViFlow and Alfa Laval. The maintenance costs may
differ due to the level of the fouling. The heat exchangers are assumed to be running 30% of
the year, therefore down-time costs are neglected. With proper maintenance the plates/tubes
would hold and function properly for at least 30 years and the costs for changing plates or
tubes are neglected. Maintenance work are assumed to be carried out once a year per heat
exchanger in order to decrease the risk of fouling and thus keep a good heat transfer between
the fluids. Table 17 shows the assumed maintenance cost per heat exchanger for each
company.
Table 17. Maintenance cost for the three heat exchanger alternatives.
Maintenance cost
Alfa Laval ViFlow GB Tank
Maintenance cost/year/HX (SEK) 59 608 5 000 8 750
Total maintenance cost of HXs for 30 years (SEK) 5 364 738 900 000 787 500
The maintenance cost for heat exchanger 1 is much higher than for heat exchanger 2 & 3.
Heat exchanger 1 is a plate heat exchanger and they are usually cheaper regarding
maintenance costs compared to shell and tube heat exchangers. The maintenance costs for the
plate heat exchanger in this case is more expensive than the shell and tube, due to plate heat
exchangers are not as suited as shell and tube heat exchangers for fluids like the oil products
Inter Terminals is handling. Heat exchanger 2 and 3 are as mentioned tubular heat exchangers
26 |
Chalmers University of Technology | and the maintenance costs are basically the same for them. Heat exchanger 2 has a little lower
cost probably due to extendable tube package.
4.4.6 Recycling
The heat exchangers are recycled after its lifetime and it is assumed that Stena Recycling AB
takes care of the recycling. After discussion with Stena Recycling, an estimated price on the
carbon steel and stainless steel were obtained (table 18). The profit of recycling the heat
exchangers is also displayed in the same table for different recycling rates.
Table 18. Price of carbon and stainless steel per ton and the profit of recycling the heat exchangers for different recycling
rates.
Recycling
Alfa Laval ViFlow GB Tank
Carbon steel (SEK/ton) - 500 500
Stainless steel (SEK/ton) 6000 - -
Weight (ton/HX) 8,8 6,4 22,3
Profit 60% recycling (SEK/HX) 31 716 1 920 6 690
Profit 80% recycling (SEK/HX) 42 288 2 560 8 921
Profit 100% recycling (SEK/HX) 52 860 3 200 11 151
Total profit 60% recycling (SEK) 95 148 5 760 20 071
Total profit 80% recycling (SEK) 126 864 7 680 26 762
Total profit 100% recycling (SEK) 158 580 9 600 33 452
The price depends on how complex the heat exchanger is to dismantle and how clean it is
when arriving to Stena Recycling. If it is not easy to dismantle and not properly cleaned,
Stena Recycling will require a sanitation fee. The price also depends on how much of the
weight is pure carbon steel or stainless steel. In this study, the total weight of the heat
exchanger is assumed to be made entirely of carbon steel or stainless steel.
4.4.7 Payback Time
When investing in new products, it is always interesting to determine the payback time.
That’s also the case for Inter Terminals, they are interested in the payback time for the heat
exchangers that have been analyzed in this assessment. Payback time can be a decisive factor
whether or not to invest in these heat exchangers in the future and has a major impact on the
decision-making. The payback time is calculated by adding the capital cost, installation cost,
maintenance cost during 30 years and the profit of recycling and dividing it with the yearly
operating cost saving. The payback time barely changes with altering recycling rate, thus will
not have an effect on the payback time.
The energy price from Göteborgs Energi is adjusted annually and affect the operating cost and
operating saving. In order to see how the payback time is affected by altering energy price
throughout the lifecycle, a sensitivity analysis was conducted on the energy price. The
sensitivity analysis takes into account both increase and decrease of the energy price by 10, 25
and 50 %. The operating costs for the current heat exchangers are fixed during this sensitivity
analysis and only the energy price regarding the new heat exchangers are changed. The
payback time is presented and analyzed in chapter 5.2.
27 |
Chalmers University of Technology | 4.4.8 Summary of the Costs for the Heat Exchangers
A summary of all the costs for the heat exchangers that were analyzed in this assessment is
presented in table 19 below. The operating costs for the current heat exchangers is also
displayed in this table. Other costs for the current heat exchangers were not available.
Table 19. All cost for the three heat exchanger alternatives and operating cost for the current ones.
Alfa Laval ViFlow GB Tank Current
Operation (SEK) (SEK) (SEK) (SEK)
Capital cost/HX 596 082 395 000 927 000 -
Total capital cost 1 788 246 2 370 000 2 781 000 -
Installation cost/HX 40 000 40 000 40 000 -
Total installation cost 120 000 120 000 120 000 -
Operating cost/year/HX 106 955 106 955 106 955 548 745
Total operating cost/year 320 864 320 864 320 864 1 646 236
Maintenance cost/year/HX 59 608 5 000 8 750 -
Total maintenance cost/year 178 825 30 000 26 250 -
Total profit 60% recycling rate 95 148 5 760 20 071 -
Total profit 80% recycling rate 126 864 7 680 26 762 -
Total profit 100% recycling rate 158 580 9 600 33 452 -
Total cost (30 years & 60% recycling rate) 16 803 763 13 010 167 13 294 356 49 387 095
Total cost (30 years & 80% recycling rate) 16 772 047 13 008 247 13 287 666 49 387 095
Total cost (30 years & 100% recycling rate) 16 740 331 13 006 327 13 280 975 49 387 095
5. Results and Analysis
5.1 Life Cycle Impact Assessment
The result from the life cycle assessment are presented in figures. It is for the three heat
exchangers and the environmental impacts are included with the different recycling rates.
Two of the heat exchangers are made of carbon steel, and one is made of stainless steel. The
results for the LCA is for the heat exchangers lifetime which is 30 years. Logically thinking,
the lifetime would affect the environmental impacts. A shorter lifetime will result in an
increased environmental impact. If the heat exchangers would have a shorter lifetime, more
heat exchangers must be manufactured. This will lead to an increased amount of metal that
needs to be extracted and more energy is needed for the steel production and heat exchanger
production. The impacts categories are global warming potential, depletion of abiotic
resources, human toxicity potential and acidification potential.
In figure 9 heat exchanger 1 has the highest impact for 60 % recycling regarding global
warming potential followed by heat exchanger 3 and 2. The global warming impact for heat
exchanger 1 has a much steeper decrease with a higher recycling rate than the other two
28 |
Chalmers University of Technology | alternatives. For 100 % recycling rate heat exchanger 3 has the biggest impact and heat
exchanger 1 has decreased to almost the same level as heat exchanger 2. For both the second
and third heat exchanger is the reduced global warming barely noticeable for the different
recycling rates. Heat exchanger 2 and 3 is made of the same material, but there is still a big
difference regarding the global warming impact. The weight of heat exchanger 2 is nearly one
fourth of heat exchanger 3. The steel production and the heat exchanger production requires
energy and the amount of energy needed depends on the weight of the heat exchanger. This
explains the difference between heat exchanger 2 and 3 regarding the global warming
potential.
For a recycling rate of 60 %, heat exchanger 1 has the highest global warming potential, even
though it does not have the highest weight. The production of steel, manufacturing of the heat
exchanger, the use phase and recycling are assumed to be the same for both carbon and
stainless steel. Therefore, the extraction of different elements and different amounts of these
elements affects the result. This means that the higher amount of chromium and nickel in
stainless steel increases the global warming impact. This can be seen in the figures 8-10,
which indicate how much the different production processes contribute to global warming
with different recycling rates. Since heat exchanger 1 is made of stainless steel, the global
warming impact decreases with an increased recycling rate. With higher recycling rates, thus
lesser extraction of virgin metals, the impact decreases. When the recycling rate increases, it
can be seen that the weight of the heat exchanger starts to influence the global warming
potential to a greater extent. As shown in figure 9, the heaviest heat exchanger has the highest
impact for 100% recycling rate followed by heat exchanger 1 and heat exchanger 2 which is
the lightest one.
Figure 7. Comparing of the heat exchangers regarding global warming potential with three different recycling rates.
Figures 8-10 shows that the use phase is the dominating process, due to the total amount of
energy that is needed to run during their lifetime. The energy is coming from three different
sources, in which all of them contributes to global warming.
29 |
Chalmers University of Technology | Figure 10. The global warming potential distribution for the process steps of the heat exchangers life cycle with 100%
recycling. SS stands for stainless steel and CB for carbon steel.
Figure 11, which describes depletion of abiotic resources looks the same as for global
warming potential. Heat exchanger 1 has the highest potential for depletion of abiotic
resources at 60 % recycling and the impact decreases with higher recycling rate and for 100 %
recycling heat exchanger 3 has the highest contribution.
Nickel and chromium is scarce metals and contains relative high amounts of stainless steel
compared to carbon steel. It is the main factor to why heat exchanger 1 has a high impact on
depletion of abiotic resources, as seen in the figures 12-14. The higher recycling rate meaning
lesser use of chromium and nickel explains the significant decrease for heat exchanger 1. As
in the case of global warming, heat exchanger 3 has the biggest impact for depletion of abiotic
resources with a recycling rate of 100 %. This is due to the weight of the heat exchanger.
Figure 11. Comparing of the heat exchangers regarding depletion of abiotic resources with three different recycling rates.
31 |
Chalmers University of Technology | Figure 14. The depletion of abiotic resources distribution for the process steps of the heat exchangers life cycle with 100%
recycling. SS stands for stainless steel and CB for carbon steel.
For a recycling rate of 60% heat exchanger 1 has the largest impact regarding human toxicity,
meanwhile heat exchanger 2 and 3 has roughly the same (figure 15). When increasing the
recycling rate to 80 %, heat exchanger 1 has a more significant decrease than the other two.
Once again it shows how the amount of chromium and nickel in stainless steel affects all the
impact categories. A tendency can be seen that for all the impact categories, an increased
recycling rate for heat exchanger 1 influence the impact categories in a positive way. Heat
exchangers 2 and 3 show a slight decrease of human toxicity with higher recycling rate. For
these two heat exchangers, the decreased impact is barely noticeable in the interval of 60-100
% recycling rate. The impact for heat exchanger 1 decreases further with 100 % recycling.
With that recycling rate all the heat exchangers have roughly the same impact regarding
human toxicity. Heat exchanger 3 will have the highest impact for all of the impacts
categories with a recycling rate of 100% due to the weight of it, which is by far the heaviest
heat exchanger.
Heat exchanger 1 has a greater impact for 60 and 80 %, than heat exchanger 2 and 3, because
in those cases the extraction of chromium and nickel are smaller. Chromium and nickel seems
to contributes more to human toxicity than the other elements. With no extraction (100%
recycling rate) of virgin metals, the heat exchangers have approximately the same impact,
which indicates that the weight of the heat exchangers barely have an effect. The kind of steel
the heat exchanger is made of is the most important factor regarding human toxicity.
33 |
Chalmers University of Technology | Figure 15. Comparing of the heat exchangers regarding human toxicity with three different recycling rates.
Chromium and nickel seem to contribute a lot regarding human toxicity except for 100%
recycling rate, especially during the extraction, as it can be seen in the figures 16-18. Since
the amount of chromium and nickel is much higher in stainless steel, heat exchanger 1
contributes much more compared to the other heat exchangers which are made of carbon
steel. With 100% recycling, heat exchanger 3 contributes more to human toxicity during the
carbon steel production due to the weight of the heat exchanger. The extraction phase has a
quite high contribution regarding human toxicity and acidification. This can be related to the
modelling in OpenLCA, in which rest of the world and global flows were used during the
extraction phase and these flows contains energy. Since the rest of the world and global
energy mix is considered to be less environmental friendly than the Swedish energy mix, it
will probably have a bigger impact for these categories.
Figure 16. The human toxicity distribution for the process steps of the heat exchangers life cycle with 60% recycling. SS
stands for stainless steel and CB for carbon steel.
34 |
Chalmers University of Technology | Figure 21. The acidification distribution for the process steps of the heat exchangers life cycle with 80% recycling. SS stands
for stainless steel and CB for carbon steel.
Figure 22. The acidification distribution for the process steps of the heat exchangers life cycle with 100% recycling. SS stands
for stainless steel and CB for carbon steel.
The figures above which shows the different impact categories, a trend can be seen. The
material that the heat exchangers are made of is crucial for all the impact categories except for
a 100% recycling rate. In that case, the weight of the heat exchangers is the main factor.
Therefore, heat exchanger 3 has the biggest impact for all of the categories. With a 100%
recycling rate, there is no extraction phase which eliminates the consideration of virgin metals
that needs to be extracted. Chromium and nickel are the main contributors for the impact
categories, since rest of the world and global energy mix were used during the extraction
phase. The high amount of chromium and nickel in stainless steel also have a big impact on
the result regarding the impact categories.
As seen in the previous figures, the use phase is the dominating process for the contribution of
the impact categories. A sensitivity analysis was conducted to see how a percentage change
37 |
Chalmers University of Technology | Table 20. Change of the impact categories for the heat exchangers (60% recycling rate) lifetime by increasing each of the
energy sources during the use phase by 20% and 40% compared to the base values. The values are in percentage and those
with a minus means a decrease.
Global Warming
Pellets Natural Gas Heating Fuel
20% 40% 20% 40% 20% 40%
Heat Exchanger 1 -1,22 -2,42 -1,07 -2,13 1,14 2,28
Heat Exchanger 2 -1,33 -2,63 -1,16 -2,3 1,24 2,48
Heat Exchanger 3 -1,25 -2,48 -1,09 -2,18 1,17 2,34
Depletion of Abiotic Resources
Pellets Natural Gas Heating Fuel
20% 40% 20% 40% 20% 40%
Heat Exchanger 1 -3,6 -7,15 0,67 1,38 0,71 1,42
Heat Exchanger 2 -3,88 -7,72 0,72 1,44 0,77 1,49
Heat Exchanger 3 -3,7 -7,35 0,68 1,41 0,73 1,4
Human Toxicity
Pellets Natural Gas Heating Fuel
20% 40% 20% 40% 20% 40%
Heat Exchanger 1 9,58 37,17 -5,37 -10,73 0,57 1,16
Heat Exchanger 2 12,89 25,8 -7,22 -14,44 0,77 1,57
Heat Exchanger 3 11,64 23,3 -6,52 -13,03 0,69 1,41
Acidification
Pellets Natural Gas Heating Fuel
20% 40% 20% 40% 20% 40%
Heat Exchanger 1 7,88 15,77 -0,95 -8,23 0,28 0,54
Heat Exchanger 2 10,44 21 -5,49 -10,94 0,34 0,71
Heat Exchanger 3 10,2 20,03 -5,21 -10,46 0,36 0,72
5.2 Life Cycle Cost
The payback time for the heat exchanger alternatives are displayed below in table 21. The
payback time considers all costs during the heat exchangers life cycle and the operating cost
savings. ViFlow´s heat exchangers have the shortest payback time of the three alternatives,
with a payback time of 2,6 years.
Table 21. Payback time for the three heat exchangers.
Payback time (years)
Alfa Laval ViFlow GB Tank
5,5 2,6 2,8
Table 22, shows the sensitivity analysis for the payback time regarding the three heat
exchangers due to a change in energy price. The payback time for the heat exchangers does
not change significantly with different energy prices. This is due to the current heat
39 |
Chalmers University of Technology | exchangers have a high operating costs. Even with an increased energy price by 50% for the
new heat exchangers, Inter Terminals would still benefit by investing into new heat
exchangers from an economic point of view. ViFlow´s heat exchanger (heat exchanger 2) has
the shortest payback time compared to the others. This sensitivity analysis indicates that these
heat exchangers are solid option for investing into, especially ViFlow´s and GB Tank´s from
an economic point of view.
Table 22. Payback time for the three heat exchangers with change in energy price.
Payback time (years)
Change in energy price Alfa Laval ViFlow GB Tank
10% 5,6 2,6 2,9
25% 5,8 2,7 3,0
50% 6,2 2,9 3,2
-10% 5,4 2,5 2,7
-25% 5,2 2,4 2,6
-50% 4,9 2,3 2,5
Investing in new heat exchangers for the three caverns would be very cost saving for Inter
Terminals when it comes to the operating cost. Over one year Inter Terminals would save 1.3
MSEK when it comes to the operating cost, by replacing the three existing heat exchangers
with new ones. Furthermore, the new heat exchangers are assumed to have a lifetime of 30
years and the cost savings for the operation cost will accumulate over that time. This results in
a cost saving of close to 40 MSEK over 30 years and with proper maintenance they would
probably function longer than that.
40 |
Chalmers University of Technology | 6. Discussion
In this chapter, thoughts will be shared regarding the results and how decisions could have
affected the result. The research questions for this assessment were the following:
Using LCA and LCC, which heat exchanger is best suited for Inter Terminals
concerning the environmental and economic aspects?
As presented in the result and analysis chapter, the best alternative from an
environmental and economic point of view is heat exchanger 2 which is ViFlow´s
suggestion. It has the shortest payback time and lowest total cost as well as the lowest
environmental impacts throughout its lifetime.
Is there a trade-off between the environmental and economic aspects?
There is no trade-off for heat exchanger 2, it is the most environmental friendly option
for each impact category and has the shortest payback time.
Heat exchanger 1 is actually over-dimensioned. It is bigger than necessary due to wrong data
were obtained from Inter Terminals at the beginning regarding the flows through the heat
exchangers and other relevant parameters related to the products. Alfa Laval did not have the
time to change their calculations regarding their alternative. Therefore, this will affect the
results, hence heat exchanger 1 probably would be smaller when designing it with the right
properties, thus lighter. This would impact both the environmental and economic aspects.
From an economic point of view, it would have a cheaper capital cost leading to a different
payback time. The environmental impacts would be slightly different, but not in a crucial way
since it is still made of a less environmentally friendly material, stainless steel.
The use of chemicals, gases and other substances during the processes were neglected due to
lack of data, if not included in the processes of ecoinvent. Neglecting these substances affects
the results, if they were included it would probably lead to a higher impact for all the impact
categories. If Environmental Product Declaration (EPD) documents were available, these
parameters would have been taken into account. For a more specific LCA, these type of data
are necessary. Also the assumption that the whole heat exchangers were made of either
stainless steel or carbon steel changes the outcome of the result from an environmental point
of view. The amount of each element in carbon steel and stainless steel can vary in many
ways, the “recipe” depends on the desired properties of the steel. Use of other ratio between
the elements for the steel will have an effect on the results for the impact categories.
There are other factors that can have an impact on the result when conducting an LCA. The
electricity mix is a major factor; in this assessment the Swedish electricity mix have been
used as much as possible for the calculations. Many companies including Alfa Laval, Viflow
and GB Tank sometimes buys complete components for the heat exchangers from other
countries. The extraction phase is also different in reality when these companies buys their
products from other countries, due to the difference in the processes in terms of energy,
transport, emissions etc. However, this assessment has its boundaries within Sweden if data is
available in ecoinvent. If not, then RoW and GLO processes and flows are used. Rest of the
world and global energy mix were used during the extraction and use phase due to lack of
data in ecoinvent. If Swedish energy mix was used, the environmental impact categories
41 |
Chalmers University of Technology | 7. Conclusion
The aim of this assessment was to conduct a life cycle assessment and life cycle costing for
new heat exchangers to help Inter Terminals with their decision-making for future investment.
The research questions were to determine which heat exchanger was the best alternative for
Inter Terminals from an environmental and economic point of view and if there were any
trade-offs between them.
By having three different heat exchangers to analyze and compare, the conclusion can be
drawn that for heat exchangers made of stainless steel, the recycling rate is more important
than for carbon steel. This is due to the amount of nickel and chromium extracted and used at
a level which is as low as possible will have a positive impact on the impact categories. The
higher recycling rate for stainless steel the better. For carbon steel, the change in the impact
categories is very low even with a very high recycling rate. When it comes to heat exchangers
made of carbon steel it is more important to reduce the weight of the heat exchanger rather
than increasing the recycling rate. This is due to the amount of energy required in the steel
production and manufacturing process. The more the heat exchanger weighs, the more energy
is needed in these two processes.
Heat exchanger 1 (Alfa Laval) has the longest payback time and also the most impact on the
environment regarding the impact categoreis for 60% and 80% recycling rate. With a
recycling rate of 100%, heat exchanger 1 has the second highest environmental impact for all
of categories.
Heat exchanger 2 (ViFlow) has the shortest payback time and is also the most
environmentally friendly regarding all of the impact categories for all three recycling rates.
Heat exchanger 3 (GB Tank) is the second best alternative regarding these aspects. However,
with a 100% recycling rate, there is a trade-off between the environmental and economic
aspects for all of the environmental impact categories. In this case heat exchanger 3 has the
second shortest payback time but has the highest impact on the environment.
From the results a conclusion can be drawn that heat exchanger 2 (ViFlow) is the best choice
regarding environmental and economic aspects and there is no trade-off between these aspects
for this heat exchanger.
During the use phase, an increased amount of pellets will decrease global warming potential
and depletion of abiotic resources. However, for human toxicity and acidification an increase
of pellets would have a higher impact for these two categories since the NO and SO
x x
emissions are higher than for natural gas. Therefore, there is a trade-off between the pellets
and natural gas when increasing the amounts of them with 20 and 40%. An increase amount
of heating fuel will lead to a higher impact for all categories. Heat exchanger 2 (ViFlow) was
the best alternative to invest into and there were no trade-offs regarding the environmental
and economic aspects.
43 |
Chalmers University of Technology | 8. Recommendations
Based on the assessment and the conclusions from this study there are some recommendations
that are of interest to Inter Terminals AB in their investment decision. Detailed data is
required from Inter Terminals regarding the flows and capacity of their pumps before
investing in new heat exchangers. The manufacturers will be able to provide customized heat
exchangers with the exact requirements.
From an environmental perspective it can be interesting to include more impact categories,
such as ecotoxicity. This can be one of the first steps that makes Inter Terminals a company
that strives towards sustainability. The environmental considerations are getting more and
more attention in Sweden but also globally, and this can make Inter Terminals to be a step
ahead of their competitors. A more detailed and specific LCA is needed for the heat
exchangers to get a higher quality of the results. This can be done by including chemicals and
substances during the different processes, since this assessment has not included those due to
lack of data. The environmental aspect of this and future assessments can be used as a
marketing tool for Inter Terminal to attract more customers in the future.
To further improve the environmental assessment a sensitivity analysis on different types of
energy mixes needs to be performed because many manufacturing company often buys
complete component for heat exchanger from different countries and this will have an effect
on the impact categories. Comparisons can also be made by looking at international
companies to see their processes and how it differs from the Swedish manufacturing
companies. Will there be a trade-off between the environmental and economic aspect in that
case? Also, how would Inter Terminals consider the trade-off? Often companies often look
for the cheapest alternatives, but in order to be considered a company that strives towards
sustainability there should be a balance between the environmental and economic aspect. This
will make Inter Terminals to stand out compared to their competitors and attract new
customers.
The economical evaluation with LCC can be improved by conducting a complete LCC. The
LCC in this assessment needs more parameters such as downtime cost and Net Present Value
(NPV). The added parameters would affect the costs throughout the lifetime of the heat
exchangers, especially the payback time. The aim from the start was to conduct a complete
LCC but due to time limitation and lack of data, it was not possible.
The methods LCA and LCC can be used for every product and not only for heat exchangers.
In the future, Inter Terminals should conduct these methods when investing in bigger
equipment’s for their facility to achieve higher environmental and economic efficiency. This
assessment provides a good base for Inter Terminals to invest in of the analyzed heat
exchangers, especially heat exchanger 2 from ViFlow. It is proved in this study that it is the
best choice for Inter Terminals from an environmental and economic point of view.
44 |
Chalmers University of Technology | Soil washing Optimisation and Assessment of the Residues of the Residues with
Focus on Copper: a Method to Treat Metal Contaminated Sites
Master of Science Thesis in Civil and Environmental Engineering
NELLY KHMILKOVSKA
Department of Civil and Environmental Engineering
Water Environment Technology
Chalmers University of Technology
ABSTRACT
Contamination of soil with toxic metals is a common and far-reaching problem of
today around the world. Metals have polluted many sites due to past and present
industrial processes, landfills and mining and pose risks for ecological systems and
human health. Only in Sweden, 80 000 contaminated sites were estimated. Currently
the most established remediation technique used for metal contaminated soils is
excavating and landfilling. This practice doesn’t solve the issue because it simply
moves the problem to a different location. On the contrary, soil washing is a
permanent alternative treatment used to remove metal contaminants from soils that
also allows for valuable metals to be recovered.
This study examined soils, which are severely polluted with copper (Cu) and other
toxic metals. In focus was developing of an enhanced soil washing method using two
leachants successively: acidic wastewater and ordinary water to dissociate toxic
metals from soil matrix. The focal point was to ensure minimal copper concentration
remained in final residues after the treatment. The changes in the soils’ texture were
studied with a view to acquire deeper understanding of the effect of acidic wastewater
on soil’s structure.
Additionally, by re-using a by-product from incineration – wastewater, this project
aimed to address the society’s increasing demand for sustainable use of materials.
Consequently, facilitating the transformation of the social attitude towards waste as a
valuable resource.
The findings from the research showed that the acidic wastewater is effective in
removing certain toxic metals from the soil matrix, in particular Cu (~90%). Still,
high leaching of Cu did not result in receiving clean enough residues to be returned
back to the original site. Nevertheless, by using the developed method to treat metal
contaminated soils, the compliance with the Swedish guidelines for non-hazardous
waste can be achieved. The final residues demonstrated ability to adsorb mercury
(Hg) from wastewater. This emphasised the importance of including pre-treatment for
wastewater prior using it for washing. The changes in soils’ structure didn’t affect
significantly its quality. The expected outcomes from further improvements on this
study is achievement of even cleaner residues, which ensures depositing the soil
residues to inert landfill or returning them back to the site.
Keywords: Soil washing, leaching, acid washing, copper, metal contaminated soil,
post-treated soil.
I |
Chalmers University of Technology | Acknowledgment
As I moved through the completion of this research work, not only I acquired deeper
knowledge in the field but most importantly met individuals without whom this
dissertation would not have been possible.
Fist, my deep gratitude goes to my thesis supervisor, Karin Karlfeldt Fedje for her
unsurpassable drive, expertise and generous guidance. My warm gratitude goes to my
supervisor and examiner, Ann-Margret Strömvall for her continuous support and
enthusiasm. I have been incredibly lucky to have such through and patient supervision
all the way. Their encouragement I will never forget.
In the laboratory I have always had the best support possible, from Mona Pålsson. Her
positive energy and generous guidance allowed my practical work to be not only
efficient but also enjoyable.
I owe a particular debt of thanks to Yevheniya Volchko for her kind help and
literature assistance that had an important contribution to this project, to Oskar Modin
for his amiable and prompt help with IC analysis and Sebastien Rauch for his
cooperative and kind help with ICP-MS analysis. I am very grateful to Haiping Lai for
finding the time and helping me with SEM session, which made a significant
difference to my project.
My sincere thanks goes to Anita Pettersson in the University of Borås, for playing a
crucial role in my quest for finding a project that I yearned to give my all dedication.
Thanks to her I had luck to work on such a project for a year.
I am indebted to Peter Therning in the University of Borås, whose forward-thinking,
percipience and trust in me allowed me to enter University of Borås in the first place.
Thank you!
I appreciate all the students and teachers in the Water Environment Technology
division for making this year with Chalmers a very special and positive experience for
me.
Nelly Khmilkovska
Gothenburg
May 2014
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
3 |
Chalmers University of Technology | Glossary
Site I Köpmannebro
Site II Björkhult
Leachant, is a liquid used in a leaching test as a leaching agent.
Eluate, is a solution (solvent and dissolved matter) produced during a leaching
process.
Soil washing, is a permanent soil treatment method used for removing metal
contaminants from soils with water or chemical solutions as a leachant.
Enhanced soil washing, is an improved soil treatment method used for removing
metal contaminants from soils. The variations may include sequential washing steps,
longer time of washing or adding washing steps with other leaching medium.
Acidic leaching, term used in this project to describe a part of batch leaching
experiment where process water was used to wash original soil.
Washing, term used in this project to describe a part of batch leaching experiment
followed after acidic leaching where soil residues were washed with Milli-Q water.
Final soil residues, is a fraction of soil that is left after original soil went through the
full batch leaching experiment.
Surface horizons, are different layers within a soil profile that are more or less
parallel at the earth's surface.
Soil structure, is a quality of a soil determined by how individual soil granules clump
or bind together and aggregate.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
4 |
Chalmers University of Technology | 1 Introduction
“Heavy metals” is a joint definition of metals and metalloids that have density greater
than 4g/!"!. Regardless, it is the fact that some of them have chemical properties
such as toxicity or ability to express poison-like quality, what causes concern
(Duruibe et al., 2007).
These toxic metals enter into the environment from natural and anthropogenic
sources. However, those that occur naturally are rarely at toxic levels. The real threat
comes from sources like mining and various industrial activities, landfills and the use
of pesticides (Duruibe et al., 2007) (USDA, 2000). Contamination of soil with toxic
metals is a common and serious problem of today. Under various circumstances
metals leach into groundwater and eventually end up in the aquifer. They can be
transported into near surface waters if metals are emitted to the run-off water, which
causes contamination of this water and consequently sediments and soil pollution.
Once metal pollutants are introduced into the environment they accumulate because of
their inability to degrade. The only exceptions are metals in organic form as for
example mercury and selenium pollutants that can be volatilized by microorganisms
(USDA, 2000). Metals in soil are sorbed on humus particles, which then passes the
pollutant along with nutrients to plants and in this way metals may enter the food
chain (Dermont et al., 2008). Moreover, high content of toxic metals have a direct
adverse affect on soil microbial health, which may have negative effect on soil
fertility (Ahmad and Ooi, 2010). Correspondingly, this leads to serious consequences
for the environment and for human health.
According to the Swedish Environmental Protection Agency (SEPA) there are around
80 000 potentially polluted sites in Sweden (Nordin, 2013).
5% 10%
Oil
10%
Polycyclic aromaQc
hydrocarbons (PAHs)
Halogenated
hydrocarbons
45%
Heavy metals
30% Other
Figure 1 Estimated distribution of pollutants in contaminated sites in Sweden (SEPA 2009).
The Pie chart (Figure 1) shows a share of different pollutants based on the top 216
sites in year 2008 (SEPA, 2009). Metal pollution represents the largest share out of all
contaminants.
To confront this environmental challenge, Swedish government formulated ”A non-
toxic environment” – an environmental quality objective along with other fifteen
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
5 |
Chalmers University of Technology | environmental objectives to be achieved by 2020 (SEPA, 2013). To address this
serious environmental threat, all polluted in Sweden sites were recognized according
to the origin and degree of pollution and the toxic effect. Currently this process is still
at the inventory phase where all of the potentially polluted sites are split into classes
(SEPA, 2013). Simultaneously, most urgent sites are treated. However, the difficulty
is that more contaminated sites arise continuously (SEPA, 2013, Ohlsson et al., 2011).
1.1 Aims and Objectives
The aim of this master’s thesis was to contribute with increased knowledge and better
understanding of acidic soil washing as a remedial method for soils contaminated with
toxic metals. A particular focus was given to Cu contaminant and evaluation of the
effect of the treatment on the soil’s structure and properties.
The aspiration of this work is to develop a soil remediation method that allows
receiving final soil residues with metals content below Swedish guidelines and near
neutral acidity.
The finding of this study should elucidate further on topic of dealing with toxic metals
in soils and methods of handling post remedial soil residues.
Specific objective were to:
• Develop an enhanced acidic washing method with all parameters for effective
washing of metals optimized.
• Investigate the effect of washing steps with water after leaching with acidic
agent; and to optimize the liquid to solid ratio (L/S) and number of sequential
steps used for this part.
• Evaluate the success of the developed method by its efficiency to leach Cu and
its ability to receive sufficiently ‘clean’ and stable soil residues.
• Evaluate and monitor changes in concentration of other toxic metals and in the
soil residues.
• Evaluate the final soil residues in terms of sensitive land use i.e. KM/MKM
and the standard leaching test SS- EN12457-3.
• Investigate the effect of the acidic wastewater on soil by comparing and
evaluating physical and chemical changes that took place after the treatment.
1.2 Limitations
The definition of the limiting factors aimed to facilitate the achievement of fair
experiments with reliable results and was described while setting the scope of the
study as well as added later as the experimental part was in progress.
The main limitations accepted and encountered in this study are listed below:
• Due to time limitation this study focused on washing of soils from deeper horizons
and excluded bark. This was true for both sites where bark was present.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
6 |
Chalmers University of Technology | 2 Literature Review
A literature study on soil as a natural valuable material, copper and current remedial
methods are presented below.
2.1 Value of Soil
Soil is a natural non-renewable resource that plays wide spectrum of vital roles that
goes far beyond the most immediately-understood ‘soil function as food production’.
According to the European Commission (European Commission, 2014), soil’s multi-
functionality can be described as “Soil performs a multitude of ecological,
economical, social and cultural functions of vital significance.” The multi-
functionality of soil is a fundamental principle of the modern understanding of soils
(Lehmann, David and Stahr, 2009). Scientific communities recognised diversity of
soil functions as (European Commission, 2014 and Volchko, 2013):
• Biomass production
• Storing, filtering and transforming substances and water
• Biodiversity pool
• Cultural environment for humans
• Source of raw materials
• Carbon pool
• Geological and archeological heritage
The quality of food and water depends on the soil’s condition. In this project, the
value of soil was explored in the context of soil’s ability to filtrate and buffer
substances - Soil as Hazard Protection - to increase the awareness of soil as a site for
adsorption, transformation and immobilization of inorganic pollutants (Lehmann,
David and Stahr, 2009).
According to Baird and Cann (2008), soils are composed of solid particles, 90% of
which have inorganic nature and the rest are organic matter and pore space. Generally,
the half of the pore space is water and the other half is air.
The silicate minerals represent the majority of the soil’s inorganic part. These
minerals composed of polymeric inorganic structures with a silicon atom as the
fundamental unit, surrounded by four oxygen atoms. Consequently, each oxygen atom
linked with another silicon unit. Some networks have aluminum ions, Al3+, instead of
silicon, Si 4+, with presence of other cations such as H+, K+, Na+, Ca2+ Mg2+ or Fe2+.
Table 1 Classification system of soil particle sizes based on the International Society of Soil Science
classification system.
Particle size
Soil type
(µm)
Clay < 2
Silt 2 - 20
Sand 20 - 2000
Gravel
> 2000
(non-soil)
On macro level the inorganic particles are the products of weathering of silicate rocks
and chemical reactions with water and acids and consist of stones, sand, silt, and clay.
The proportion of these components determines soil’s texture.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
8 |
Chalmers University of Technology | Depending on the size of these particles the soil can be classified into different types,
see Table 1. Sand classifies further into fine sand (20-200µm) and coarse sand (200-
2000µm) (Baird and Cann, 2008).
The total surface area of a clay particle per gram is thousands of times larger than that
of a silt or a sand that’s why they act as colloids upon contact with water.
Accordingly, most significant mechanisms in soil such as retaining nutrients or
binding organic matter happen on the surface of colloidal clay particles (Baird and
Cann, 2008). The fractions of different sizes form soil’s skeleton and affects its
physical, chemical and biological characteristics.
The organic part consists primarily of humus - partly decomposed photosynthetic
plant material (Baird and Cann, 2008). The organic matter is an important component
of soil and it represents 1-6% of the soil’s structure depending on a soil type. It
characterized by dark colour and primarily consists of humus. The humus content has
a direct effect on physical, chemical and biological qualities of the soil. In this project,
the terms humus and organic content used interchangeably implying the same
meaning.
2.2 Copper
The use of Cu metal accounts for at least 10,000 years. It is widely used, especially
for electrical wiring in telecommunications, building and technology sector. In
addition, Cu used to produce brass and bronze alloys, expanding further the spectrum
of its use.
The versatility of Cu is a result of its unique qualities, such as high toughness and
ductility. Moreover, Cu is only second to silver in conductivity, making it an
exceptional conductor of heat and electricity (European Copper Institute, 2014).
Although reserves of Cu are still sizeable, there are too many variable to conclude
confidently that Cu source is infinite. Such as scientists neither know exactly how
much of Cu sources there are in Earth’s crust nor its exact locations. The global Cu
consumption reached 19.8 million tonnes per annum in 2011 and expected to rise in
the future. Moreover, with China’s and India’s increasing demands for metals, Cu
supply becomes increasingly constrained which leads to increased Cu prices, see
Figure 2 (London metal exchange, 2014 and InvestmentMine, 2014).
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
9 |
Chalmers University of Technology | Figure 2 The change of prices for copper commodity over period of 25 years (InvestMine).
According to InvestmentMine the price for copper commodity is 4,781 EUR/t (2014-
01-08). Because Europe is only the third in the world in production of Cu (19%) after
Asia and America, it is a big advantage to have technology that allows reuse of Cu
extracted from soil and other polluted materials. This goes hand in hand with is the
main motivations for this project - to remove Cu pollutants permanently from soil
matrix. Another anticipated advantage of the developed method is an opportunity to
re-introduced Cu back into society (London metal exchange, 2014; Karlfeldt Fedje et
al., 2013).
Copper as a pollutant that acts similarly to other toxic metals: it bounds to water
sediments or soil particles, not able to biodegrade and accumulates in the ecosystem.
For humans Cu does not present a serious danger unless repeated exposure to
extremely high Cu concentration occurred for a long time. However, Cu has a
negative environmental effect on biological life in the polluted areas (Fenglian and Qi,
2010). Moreover, the consequences of traditional Cu mining have both immediate and
chronic effects on the landscape, waterways and growth of vegetation in the
surrounding areas with production of toxic waste as a side stream (Faculty Virginia,
2009). So if an alternative way to acquire Cu should be discovered, this will have far-
reaching and positive effects.
2.3 Remediation Methods
Different types of pollutants require different treatments. The targeted pollutants in
this report are metals; accordingly, the investigated remediation methods are dealing
with the same group of contaminants.
Choice of a treatment procedure is influenced by several factors and complicated by:
heterogeneous distribution of contaminants; chemical and physical variations of metal
forms within the soil matrix; the fact that metals are non-degradable (Dermont,
Bergeron and Mercier, 2008). Consequently, the excavation and landfilling is by far
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
10 |
Chalmers University of Technology | the most common choice of dealing with contaminated soils. It is an established
method, providing a fast result, not requiring special equipment (except for
transportation) or specialists (Ohlsson 2011). However, this method does not provide
a sustainable solution to treat soils. Therefore by landfilling heavily contaminated
soils the problem is moved from one location to another, while valuable metals that
otherwise could have been recovered are buried again. Additionally, an increasing
interest in more sustainable remediation treatments has been observed globally. It
demands for cost effective technologies that allows permanent solutions to the
problem (Shammas, 2009).
Currently there are two ways to approach soil remediation: immobilisation/isolation
(I/I) or extraction of the metal contaminants (Figure 3). Additionally, depending on
the extraction feasibility of a particular metal, the size of a contaminated site and
available infrastructure the choice between in situ (on site) and ex situ (off site)
treatments is made (Dermont, Bergeron and Mercier, 2008). When immobilisation
methods are used, the main aim is to stabilise metals and minimise their leaching.
However, no immobilization treatment technology is permanently effective because
metal contaminants are still remain in the soil matrix and therefore may leach/release
under various changing conditions. Such as one of the common techniques –
stabilization - according to Dermont, Bergeron and Mercier (2008) the behavior of
stabilized soils in a long-term perspective is not sufficiently researched. Moreover, the
area left after the soil has been excavated (ex situ) has to be refilled with clean soil.
Figure 3 Schematic diagram showing some of the existing remediation technologies for metal-contaminated
soils.
For this reason, extraction methods that aim at removing metals from the soil matrix
and ultimately decontaminating the site, may offer a more favorable solution.
Nevertheless, given the highly heterogeneous nature of metal contamination in soils,
often metals extraction is difficult. A lack of economic viability is another important
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
11 |
Chalmers University of Technology | factor. Such as well-established technique of physical separation may allow fast
solution with immediate reduction in contaminated volume. However, this method is
difficult to use for high clay or humic content soils (Mulligan, Yong and Gibbs,
2000). Phytoextraction is an in situ technique that uses plants to extract metals from
the soil. By using this procedure, a large area can be treated without excavation.
However, the process duration is long and limited by depth of the root zone and
harvesting routine (Dermont, Bergeron and Mercier, 2008).
Soil washing is an ex situ technique that uses an extracting aqueous agent such as
acid, base or chelating agents to extract metal from soil. It allows to remove
permanently metals from the soil’s matrix with a potential to recover it. However,
soils with high humic and clay content are difficult to treat, while chemical leaching
agent can be expensive and possess hazardous qualities (Dermont, Bergeron and
Mercier, 2007; Abumaizar and Smith, 1999). Looking to address the above
challenges, this project further explored and expanded the possibility to permanently
remove metals from soil by using soil washing with acidic wastewater. Moreover,
through using a waste product instead of expensive reagents the method aims to
improve economic viability.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
12 |
Chalmers University of Technology | 3 Studied sites
3.1 Köpmannebro
There is a contaminated site in Melleruds municipality, called Köpmannebro, see
Figure 4. The contamination was caused by wood processing industry for utility poles
in the early 1900s. According to the past used technology, blue vitrol was injected
into the trees. The blue vitriol consists of one Copper (II) sulfate molecule that is
crystalline bonded to five water molecules [CuSO ⋅5H O]. While the operation, spills
4 2
were occurring which was the initial source of the contamination. The main cause,
however, was due to contaminated bark and branches that were allowed to lay a side
causing continuous pollution over a period of time. In addition to Cu pollution, in
some spots a raised level of lead was detected (Kemakta, 2012).
The former industry left behind a highly contaminated site of 8000m2 with 70% of the
samples taken at the site showing Cu concentration associated with toxic waste. It was
estimated that within the area used for wood impregnation, over 30 tonnes of Cu is
enclosed. Soil, peat and bark all carry Cu contaminant with bark being the most
affected. Consequently, because all of the soil layers, ground water and sediments are
polluted to various degrees, the lack of vegetation can be observed in the area, see
Figure 4.
Figure 4 The contaminated area in Köpmannebro.
In some part of the site, the bark horizon reaches 1m thick. However, in this study, the
focus was given to soil below the bark’s horizon. The technologies of landfilling or
solidification were suggested by Kemakta Konsult AB (2012). For convenience, in
this study Köpmannebro will be referred as Site I.
3.2 Björkhult
The second site investigated in this project is Björkhult, situated on the south shore of
the lake Verveln, 15km south of Kisa in Kinda municipality (Figure 5). From 1916
and for nearly 30 years there was a wood-processing factory for telephone poles. The
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
13 |
Chalmers University of Technology | technology used a 1,5 – 2% Cu solution for impregnating trees-trunks with bark
removed following the impregnation. The removed bark and small branches were left
lying around which was the cause for the area of about 7300m2 to be heavily
contaminated with Cu. Raised concentration of Cu detected as deep as 4 meters.
According to WSP Environmental (2010) the Cu contamination was detected down to
at least 3m with a significant part occurring below the groundwater level. The bark
horizon is the most affected and situated mostly in the upper 0,5m of the soil’s profile.
The total amount of Cu within the area is estimated to be around 25 tonnes (WSP
Environmental, 2010).
Figure 5 The contaminated area in Björkhult.
According to Eriksson and Johansson (2013) during sampling, three well-defined soil
horizons could be identified and described as: 0-10cm - sandy soil, 10-30cm -
incomplete degraded bark layer and below 30cm - red soil (finer-grained than the top
layer), see Figure 6. In this study, this site will be referred as Site II.
Figure 6 Soil horizons observed while samples collection, Björkhult.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
14 |
Chalmers University of Technology | 4 Criteria for contaminated material
In this chapter, the order and methods to evaluate studied soils is described. The
original soil before washing and final residues received after the treatment were first
evaluated against KM/MKM guidelines to investigate the possibility to return the soils
back to the site. If, the soil did not comply with the KM/MKM guidelines, the next
step was to evaluate the soil residues as waste by performing the Standard Leaching
Test to determine the suitable type of landfill.
4.1 KM and MKM guidelines
First guidelines applied to evaluate pre- and post remedial soils in this project were
KM/MKM, developed by Swedish Environmental Protection Agency for soils and
build upon the general land use. KM – stands for sensitive land use and MKM – less
sensitive land use (SEPA 2009c, report 5976). The differences between the above
land uses are explained in Table 2.
Table 2 Definitions for protected objectives under general land use, KM and MKM guidelines (SEPA 2009c).
Protected Objectives KM – sensitive land use MKM – less sensitive land
use
People presence in the Full-time stay Part-time stay. Elderly and
area children- occasional to none.
Soil environment in Protection of soil ecological Limited protection of soil
the area function. ecological function.
Groundwater Protected within and Protected in a downstream
adjacent to the protected distance of 200m.
area.
Surface water Protection of surface water Protection of surface water
protection of aquatic protection of aquatic
organisms. organisms.
The generic guideline values of metals concentrations are designed to assure a
protection for people living on or visiting the site and were used to evaluate soil
samples in this study, see Table 3.
Table 3 Generic guideline values showing KM and MKM guidelines according to Swedish EPA for metals.
Substance KM mg/kg MKM mg/kg
As 10 25
Pb 50 400
Ba 200 300
Cr (total) 80 150
Cd 0.5 15
Co 15 35
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
15 |
Chalmers University of Technology | Cu 80 200
Sb 12 30
Zn 250 500
Hg 0,25 2,5
Mo 40 100
Ni 40 120
V 100 200
Thus, the requirements to the soil quality according to the KM/MKM-guidelines
depend on the soil applications, existed and likely activities on the particular site and
the age groups of people that potentially will be exposed to contaminants.
4.2 Swedish Standard Leaching Test SS-EN-12457-3
Second stage in evaluating of the pre-and post remedial soils if they had
concentrations of metals above the KM or MKM guidelines was to do a downscaled
leaching test SS – EN 12457-3 (SIS, 2003). This stage treats soil as a waste and called
waste characterisation. This is a compliance test applicable to use on a material that
have at least 95% (mass) grain size less than 4mm. Appropriately, the soils from both
sites were suitable to use.
This leaching test assessing the mechanism of release of soluble pollutants when
granular waste material is in contact with water. This mechanism predicts the
potential risk to the environment if the soil is re-used or disposed to a landfill. Thus,
the purpose of this test is to predict stability of the soil.
The released soluble constituents were measured for metals concentrations and
analysed. To interpret the results, the guidelines developed by SEPA were applied.
According to SEPA report (NFS 2004:10), the waste can be classified as inert, non-
hazardous and hazardous. For each of the classes the table is given with metals’ and
organic materials’ limits stated. Based on the results, the soil can be given one of the
three classes and determine whether it can be deposited in a landfill for non-hazardous
waste.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
16 |
Chalmers University of Technology | 5 Methods and Materials
The project began with literature study, followed by laboratory work. The literature
research was done to get an overview of the current remediation techniques of metal
contaminated soils (MCS). The outcomes of the literature study were: determination
of suitable values for variables used in the leaching experiments and the development
of a plan for experimental part of the project.
The laboratory part of the project commenced when the appropriate experimental data
was collected and made into a final flow chart helping to navigate throughout the
practical work.
5.1 Preparation
Before the experimental part was commenced some additional preparation had to be
made. The description of these steps is given below.
Sampling
Soil samples from the two sites were collected in 2012 from the areas marked as Cu
‘hot spots’ in Köpmannebro (Site I) and Björkhult (Site II) (Kemakta, 2012 and
Arnér, 2011). The soil was collected at specific depths by means of stainless steel
shovels. The samples were then stored in PP-bottles at 4°C for around a year.
Sample preparation
In order to prepare a single representative sample for Site I and Site II, soils from
several ‘hot spots’ were mixed together. For Site I, a mortar was used to break dry
clay-like soil agglomerates but avoiding intensive grinding.
The soil samples collected from depths between 50–80cm was mixed together with
those collected deeper than 100cm to make mixed sample from Site I. For Site II, soils
taken between 20–50cm and 40–60cm were mixed. The final mixture was screened
and materials larger than 2cm removed.
The samples were dried in an oven (Memmert U15) at 104°C, until their weights were
constant. The first two hours of the drying procedure the samples from both sites were
gently stirred several times to avoid formation of a solid cluster. After the samples
were completely dried, they were cooled and stored in desiccators.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
17 |
Chalmers University of Technology | For both sites the sieve was used in the same manner except that the soil sample from
Site I had to be mortared gently to break agglomerated clay particles prior the sieving.
The sieve with a soil sample was shaken for 2 minutes and then allowed to stand for a
minute before opening. This was done to ensure that smaller fractions were settled.
The fraction size distribution (w-%) was calculated for the each site.
To find out what fractions of the studied soils are the most contaminated, each soil
fraction was leached separately and handled according to the schematics in Figure 7.
The eluates produced after the leaching procedure were analysed for Cu concentration
by semi-quantitative HACH method. Then, Cu concentration in the original samples,
of each fraction size was estimated based on the efficiency of 90% and depending on
the Cu concentration measured in the corresponding eluates. The assumption of the
90% leaching efficiency was made based on the early studies by Karlfeldt Fedje et al.
(2013). The latter showed that the acidic process water could leach ≥ 90% of Cu. To
receive more accurate data some eluates were also analysed by ICP-AES.
Liquid to solid ratio
Fixed parameters used:
• Optimum process water
• Site I – not sieved, Site II – sieved
• Triplicates per tested L/S-value
Three L/S values were tested: 8, 9 and 10. To find the optimum L/S (L/S ),
opt
triplicates were leached for each L/S-value to acquire a reliable data. The soil samples
from both sites were leached according to Figure 7. The semi-quantitative HACH
method was used for the chemical analysis of Cu concentration in all eluates. The
L/S-value that leached a highest concentration of Cu was taken as the L/S and used
opt
in final batch leaching procedure.
Leaching procedure for optimisation part
To perform the optimisation part the following leaching procedure was used:
Pre-mixed and dried soil samples from both sites of 4g was placed in 50 ml PP-bottles
and marked. The optimum process water was added to reach a specific L/S. The
samples were leached for 30min on shaking table Julabo SW-20C with 140rpm.
Immediately after the shaking table, a centrifuge Sigma 4-16 was used to separate the
liquid from the solid part at 3000G for 15minutes. After the centrifugation was
complete, the liquid part was decanted.
The residence time and settings for defining the intensity of shaking on the reciprocal
table were taken from the previous studies by Eriksson and Johansson (2013) and
Karlfeldt Fedje et al. (2013). The solid part was discarded at this stage of the
experiment, focusing on the supernatant. The liquids were filtered with the acid-proof
micro-glass fibre filters (Munktell), followed by the analysis with HACH or storing in
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
20 |
Chalmers University of Technology | the fridge at 4°C. The schematics of the leaching procedure presented in Figure 7.
Soil
sample
4g + process water
(specific L/S)
Shaking table 30min at 140 rpm
Centrifuge 15min at 3000G
HACH
and/or
Solid residue Filtration
ICP-MS
Figure 7 Flowchart showing the method used for leaching procedure and applied in optimisation part of the
laboratory work.
5.3 Batch leaching
The aim of the batch experiment was to test the washing procedure with optimised
parameters on a larger amount of soil – scaling up. The amount of soil used for each
site was 100g. For Site I the whole sample was used while for Site II only three
smallest fractions (<0.25mm - <0.125mm) were taken. Thus, the soil had to be sieved
to gather 100g of the target fractions.
Simultaneously, the developed acidic washing method was expanded further by
including steps where Milli-Q was added. The latter aimed to release weakly adsorbed
Cu2+ ions and to raise the pH of the soil residues. The optimisation of number of such
steps as well as L/S ratio applied was done concurrently.
For simplicity, the part in the batch leaching experiment when process water was used
as a leaching agent referred as acidic leaching, while the following part of the
experiment when Milli-Q water was used referred as washing or washing part.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
21 |
Chalmers University of Technology | Acidic leaching
The dried soil sample of 100g was separated into two equal parts and placed in two
500ml PP-bottles. To each sample the optimum process water was added with the
L/S (L/S = 8). Further, the samples were treated according to the same method as
opt
described in Figure 8. The solid residues received after the centrifuging were
separated into two parts (≈75/25%). The smaller part was dried in the oven (Memmert
U15 at 80°C) and stored in a desiccator until further analysis. The larger part was
further divided into two equal parts that were simultaneously used in the washing
experiment that followed directly after the acidic leaching.
Washing
The optimisation of L/S ratio and the optimum number of washing steps was
performed in duplicates. The experiment was done on the same day when duplicates
were formed from freshly centrifuged solid part (≈75%) that was left after the acidic
leaching. With a pipette, Milli-Q was added sequentially to each of the duplicates
placed into the plastic holder on a filtering apparatus. Each step had a specific L/S
ratio. The L/S values were estimated based on a centrifuge-dry soil; the weight change
was traced through the each stage of the experiment. After adding each water portion
to the residue, it was left to stand for 2 min before starting a vacuum filtration. This
was done to allow the Milli-Q water to flow freely through the larger pores within the
soil matrix before forcing it downwards by the vacuum suction.
For each L/S-value added, an eluate was collected separately. The values of L/S that
were used in the laboratory work are given in Figure 8. In addition, one of the
duplicates in the soil residue from Site II was tested for L/S=2.1. The residues that
were left after completing of all sequential washing steps were collected and dried in
the oven (Memmert U15) at 80°C until their weight was stabilized. The dry residues
were stored in desiccators until further analyses. The full procedure of batch
experiment is shown in Figure 8.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
22 |
Chalmers University of Technology | 100g soil + 800ml p.w.
Process water added
(leaching part)
Shaking for 30min
Supernatant Solid residue (≈25%)
Ext. lab Dried/Stored
Centrifuge 15min
Milli-Q water added
(washing)
Solid residue (≈75%)
Duplicates
L/S = 0.1
Residue+ Residue
L/S = 0.5
Ex. filter +filter. Ex.
lab equipme equipme lab
nt L/S = 0.9 nt
L/S = 1.3
**L/S=2.1
Residue/ Residue/
Dried/ Dried/
External External
lab lab
*L/S=1.7 was used for all soil samples, but not showed graphically to simplify the diagram.
** L/S=2.1 was used only for soil sample from Site II in one of the duplicates.
Figure 8 Flowchart showing the full procedure of batch leaching followed by washing steps optimisation.
5.4 Characteristics of soil residues after batch leaching
It is expected that soil would change chemically and physically after the acidic
leaching. To give an account of these changes, the soil residues were analysed.
Selected original and corresponding final residue samples were sent to an external
laboratory to be tested for: pH, humus content, clay content and soil texture. The
results were compared and used in soil function evaluation according to TUSEC
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
23 |
Chalmers University of Technology | manual. In addition, scanning electron microscope (SEM) was used to get
supplementary information about internal particle structures and solid phase
speciation of metals in the soil matrix.
Soil as Filter and Buffer for Heavy Metals
The Institute of Soil Science and Land Evaluation developed TUSEC manual - a tool
for schematic assessment of natural and anthropogenic soils (Lehmann, David and
Stahr, 2009). The manual enables soil evaluation and categorization independently of
specific pedological methodologies. With help of this methodology a wide selection
of soil functions can be evaluated as well as individual soil performances.
According to Senate Department for Urban Development and the Environment
(2009), buffering function expresses soils ability to impede movement of toxic metals
in the ecosystemic material flow. The soils ability to eliminate them completely from
this cycle describes by soil’s filtration function.
To answer the question: how efficient is the studied soil in buffering and filtering of
toxic metals, it was chosen to assess it’s following soil function:
Soil as Filter and Buffer for Heavy Metals (STOFIT3’B)
This evaluation performed on original soil samples and final soil residues and allowed
to rate the soils as follows: “very low” – 5, “low” – 4, “intermediate” – 3, “high” – 2,
“very high” – 1.
The TUSEC manual offers two procedures: A-procedure (not chosen) - demands for
high quality of available data of every single horizon; B-procedure (chosen) - based
on soil data that does not take into account every individual soil horizon (Lehmann,
David and Stahr, 2009). Moreover, because B-procedure does not require a
standardized data collection, it is easier to apply it due to time constrictions. The
results of B-Procedure should be used for orientating evaluation.
The following input parameters for both sites were used:
1. Clay content
2. Humus content
3. pH-value
4. Information on average groundwater level
The procedure described in TUSEC manual is in step-order. The information
regarding groundwater level was taken from Kemakta and WSP Environmental
reports for the Site I and II respectively (Kemakta, 2012 and WSP Environmental,
2010).
As a result, a score of 1 to 5 was received.
pH measurement of soil samples
In this study, the method for measuring pH in the laboratory was done according to
ISO 10390: 1994 (Bergil and Bydén, 1995) using a WTW pH-electrode SenTix 41-3
with a WTW Multi 35i.
The external pH analysis of the original and post-remedial soil samples were done
according to the SS – ISO 10 390 method with uncertainty of < 0.4% (for selected
samples). These results were compared to the internal laboratory readings (quality
control) and the average of two analyses was given as the final answer.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
24 |
Chalmers University of Technology | Humus content
The soil was pre-treated for analyses in agreement with SS – ISO 11 464 standard,
ed.1 regulations. The humus content was analysed according to KLK 1965:1 with
uncertainty <5%. The received results were used as one of the input parameters in the
evaluation of the soil function as Filter and Buffer for Heavy Metals as well as in
determination of a soil type.
Clay content
Clay content analysis detects the weight % of clay minerals, formed by weathering of
the silicate minerals from rock. The clay content was measured according to SS – ISO
11277:2009 standard with uncertainty <10%. The results were used to disclose soil
texture and served as one of the input parameters in evaluating soil function as Filter
and Buffer for Heavy Metals.
Soil texture
The soil texture was evaluated as stated in SS - ISO 11277:2009 standard with
uncertainty of <10% by an external laboratory based on samples of 60g taken from
each site. A joint name was given to each of the studied soil, based on the amount (%)
of the humus and its soil texture, see Table 14.
SEM - Scanning Electron Microscope
Scanning electron microscope coupled with energy dispersive x-ray EDX was used to
produce magnified images of the soil samples. The visual assessment was done to
compare the external morphology and orientation of granules making up the sample
(Geochemical Instrumentation and Analysis, 2013).
The following samples were assessed:
Site I - original soil and final residues after full washing procedure.
Site II - original sieved soil and final residues.
The following magnifications were used on each sample before taking a photograph:
200x, 500x, 1000x and 2000x. Then, on two randomly chosen sections of different
magnifications, spot analyses were performed (Figure 9). The spot analysis revealed
chemical composition in a particular spot. To receive a better sample representation,
10 spot analyses were performed in total for each analysed sample.
Section 1 Five spot
Magnification
analyses
on a sample:
Results
200x, 500x analyses
1000x, 2000x Five spot
Section 2
analyses
Figure 9 Sc hematics showing steps followed for each soil sample in SEM analysis.
Swedish Standard Leaching Test
The dried pre- and post- washed soil samples of 2g were:
• Leached with Milli-Q for 6 h with L/S = 2
• Followed by leaching with Milli-Q for 18 h with L/S = 8.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
25 |
Chalmers University of Technology | The standard leaching procedure and equipment were used, see Figure 7. Each
sample’s liquid part was decanted, its volume measured and filtered. The resulted four
eluates were marked and stored at 4°C awaiting analysis with ICP-AES.
The following substances were not analysed and therefore excluded from evaluation
at this stage: Hg, Se, Sb and Cl. The organic pollutants were not studied in this project
but they are present in the NFS 2004:10 as criteria for waste evaluation.
5.5 Analytical Methods
HACH
The HACH is a photometric instrument designed to test the quality of water or liquid
solutions. In this study it was used to receive quick internal measurements on Cu
concentration. However, the detected Cu amounts should be seen as semi-quantitative
rather than exact values. Consequently, HACH results were used for comparing
between samples in order to determine trends.
The standard HACH Method 8506 was applied. Furthermore, the Bicinchoninate
methodology was chosen (HACH manual, 2009). Before analysis the following
adjustments had to be performed for most eluates:
• Dilution, since all elutes were extremely concentrated with Cu
• pH adjustment between 4-6 by adding pH 8M KOH.
External analyses
To receive quantitative data, selected solid and liquid samples were sent to external
certified laboratories. Their services were used to measure metals concentration in
eluates, process water, original soil samples and final soil residues.
Prior to the analysis, all eluates were filtered. If a sample had pH above 2, it was
acidified by using 1% HNO in proportion 1ml HNO to 100 ml of a sample
3 3
(concentrated HNO ). Some of the samples were diluted as well.
3
All liquid samples were analysed using ICP-AES according to EPA methods 200.7.
To analyse Hg and Se the ICP-AFS were used, according to method SS-EN ISO
17852. While, Sb was analysed by ICP-SFMS instrument according to US EPA
200.8 method.
Analyses of solid samples were done according to EPA methods: (modified) 200.7
with ICP-AES instrument and 200.8 with ICP-SFMS instrument. To analyse As, Cd,
Cu, Co, Hg, Ni, Pb, W, Sb, S, Se and Zn the samples were dried at 50 ° C and the
element concentrations were TS-corrected to 105 ° C. Microwave-assisted digestion
was used with 5 ml of concentrated nitric acid and 0.5ml H O . For Sn, digestion with
2 2
reverse Aqua Regia was used. For other elements the following method steps were
used: 0.1g dried sample was melted with 0.4g lithium metaborate (LiBO ) and
2
subsequently dissolved in dilute nitric acid (HNO ).
3
To test the sample with the ICP-MS instrument it was first converted into an aerosol
and injected into the OCP-torch where the sample evaporated. Through collision with
electrons the sample atomized and ionized. The analysis detected the atom ions based
on their M/Z-ratio (Thomas, 2004).
The original soil samples and final soil residues that were left after the fully optimised
batch experiment were tested for pH, soil texture, humus and clay contents in a
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
26 |
Chalmers University of Technology | 6 Results and Discussion
The results presented below follow the order in which laboratory work was
performed. At the end, a prospect of the developed method from point of view of
process water availability is discussed.
6.1 Characteristics of original soil
Before experiments could be commenced, the soils from both sites were studied and
evaluated. Some of the initial information was taken from the preceding reports and
studies (Kemakta, 2012 and WSP Environmental; Eriksson and Johansson, 2013). The
parameters presented below were studied and evaluated in the laboratory.
pH
The pH results acquired via external analysing and internal laboratory results are
summarised in Table 5.
Table 5 Comparison of pH measurements after internal and external analyses for soils from Site I and Site II.
Site I Site II
Analysis
(Köpmannebro) (Björkhult)
Internal 5.0 5.2
External 4.6 4.9
It could be seen that the results show a similar pattern with the data received from the
internal laboratory giving slightly higher values. This distinction can be explained by
different methods that were used, see Chapter 5.4. The soil’s heterogeneity is likely to
contribute to this variation as well. The average of the internal and external reading
for Site I=4.8 and for Site II=5. These results were taken into further soil evaluations.
The pH of the original samples in both sites does not differ significantly. According to
Moody (2006) the majority of soils have pH between 3.5 and 10 with those labelled
‘strongly acidic’ below 5.5. However, 70% of soils in Sweden belong to Podzol type
with natural pH around 4. Thus, with the geographical considerations, the soils from
Site I and Site II can be considered to have a normal pH (MarkInfo, 2007).
Soil texture
The original soil from Site I had a clay-like nature, while the soil from Site II can be
characterised as sandy, see Figure 10.
The result of sieving showed that soil from Site I consists mostly of fine sand
(0.125mm ≥ 34% < 0.25mm) while the largest representative fraction in Site II was
≥1mm - coarse sand, see Table 6. Also the percentage of silt-clay fraction (<0.125)
differs significantly between two sites, with Site I showing a much higher
representation.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
28 |
Chalmers University of Technology | Table 7 Comparison of metals concentration in original soils from Site I and Site II, analysed by ICP-AES.
Substance Site I Site II KM mg/kg MKM
mg/kg original original TS mg/kg TS
As 0.4 0.6 10 25
Ba 526 777 200 300
Cd 0.0 0.1 0.5 15
Co 2.1 1.1 15 35
Cr 25.7 27.6 80 150
Cu 2390 2200 80 200
Hg <0.04 <0.04 0.25 2.5
Ni 3.9 2.4 40 120
Pb 8.2 5.9 50 400
V 46.1 23.5 100 200
Zn 13.0 13.2 250 500
It is clear that the original soils from both sites are heavily contaminated with Cu. The
concentration of Cu in both sites higher than MKM guideline: 12 times higher for Site
I and 11 times higher for Site II. The metal analysis did not confirm that Site I have a
high content of Pb as it was stated in the Kemakta report (2012). This can be
explained by heterogeneous distribution of Pb within the soil matrix as well as within
the contaminated site itself.
6.2 Optimization Part
Process water optimization
After leaching the four soil samples from Site I with process water of different
dilution ratios, four corresponding eluates were produced and analysed for Cu
concentration, see Table 8.
Table 8 Copper concentrations and pH analysed in eluates produced by washing soil from Site I with process
waters with L/S=10 and different dilution ratios.
Dilution ratio (acid/Milli-Q) pH Cu mg/L
100 0.1 8.0
75/25 0.3 8.3
50/50 0.4 5.8
35/65 0.6 6.4
The original process water and the solution of 75% pure process water mixed with
25% of Milli-Q water (75/25) leached similar amounts of Cu, while further dilution of
process water resulted in less efficient release of Cu. Consequently, (75/25) process
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
30 |
Chalmers University of Technology | water was used in all further experiments. By using less concentrated process water,
two positive outcomes can be expected:
• Lower negative effect on the soil matrix
• Less process water can wash more soil
For future experiments it is desirable to repeat the above optimization of process
water in triplicates to ensure more reliable results. Moreover, because chemical
composition of process water varies significantly it is advisable to do the dilution
optimization for every new process water taken from the industry with intention to be
used as a leaching agent.
Sieving optimisation
To evaluate suitability of physical separation as pre-treatment method the soils from
both sites were sieved. The results showing particle size distribution, correlated to Cu
leachability in soils from both sites are presented in Table 9.
All fractions in Site I showed a similar concentration of Cu leached, thus leaching
ability is independent of fraction size. According to Dermont, Bergeron and Mercier
(2008) the fact that Cu pollutant has no specific location may indicate that it is mainly
in a particulate form. With assumed 90% leaching efficiency of Cu, it was estimated
that the concentration of Cu in all fractions was above or close to MKM guideline.
Therefore, all fractions were equally contaminated and consequently all soil fractions
had to be washed (Figure 11).
Table 9 Copper concentrations analysed by HACH after leaching each soil fraction in Site I and Site II
separately, mg/L.
Sieve size, mm Cu leached, mg/L
Site I Site II
≥ 1 10 6
1 - 0.5 8.7 8.9
<0.5 - 0.25 8.5 12
<0.25 - 0.125 8 29
< 0.125 13 49
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
31 |
Chalmers University of Technology | Cu concentration in each fraction estimated
with 90% efficiency
4000
3500
3000
2500
2000
Site I
1500
Site II
1000
500
MKM
0
<0.125 0.25 -‐ 0.125 0.5 -‐ 0.25 1 -‐ 0.5 >1
Soil fractions (mm)
Figure 11 Predicted Cu concentrations in each soil fraction (original soils in Site I and Site II), assuming 90%
leaching efficiency. Based on HACH analysis.
It is evident, that in Site II the smallest three fractions are the most polluted with Cu,
see Table 9. The fraction < 0.125mm leached 28 times more Cu than fraction >1mm.
Thus, the trend is: increased Cu leaching with decreased grain size (Figure 11). With
assumed leaching efficiency of 90%, the three smallest fractions showed Cu pollution
much higher than the MKM guideline. The mixture of the two largest fractions (≥
1mm and 1 - 0.5mm) represents 78% of the total soil sample but estimated to have 26
times lower concentration of Cu compared to the < 0.125mm fraction. This means
that by excluding it, far less soil has to be treated. Based on that, for Site II it was
decided to carry on physical pre-treatment further i.e. using only the smallest three
fractions for enhanced soil washing.
L/S optimization
To optimise L/S ratio, soil samples from both sites were leached with optimised acidic
process water of different L/S-values. Three eluates were produced and analysed for
Cu, see Table 10.
Table 10 Semi-quantitative values of copper concentration analysed by HACH in eluates produced with
different L/S-values.
Cu leached, mg/L
L/S
Site I Site II
8 13 18
9 12 16
10 11 15
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
32
)gk/gm(
uC |
Chalmers University of Technology | It can be seen that using L/S=8 yielded the highest Cu leaching from both sites.
Although there is no significant difference between Cu leached for each L/S-value, a
clear correlation can be observed: the lower the L/S-value the higher the concentration
of Cu in the eluates. Based on this, L/S=8 was accepted as optimum (L/S ). The
opt
possible explanation for lower L/S ratio to be more effective in leaching Cu is that in
a smaller volume more soil particles more frequently come in contact during a set
leaching time. This may result in grain over grain mechanical brushing with more
active sites being activated and available.
6.3 Batch experiment
In the acidic leaching part of the batch experiment all parameters were optimised:
L/S=8; dilution ratio process water to water = 75:25; Site I – no physical pre-
treatment (sieving), Site II – physical pre-treatment included.
However, the washing part was optimised concurrently i. e. the L/S and number of
opt
steps.
Washing optimisation
The results in Table 11 show how much Cu was leached when Milli-Q water of
different L/S-values was sequentially added stepwise to the same soil residue i.e. after
the acidic leaching part in the batch experiment.
Table 11 Concentrations of Cu, analysed by ICP-AES in eluates produced in the washing part of the batch
experiment when Milli-Q water of various L/S-values was sequentially added.
Parameters – washing with
Cu leached, mg/L
Milli-Q water
Step number L/S Site I, mg/L Site II, mg/L
1 0.1 3.6 9.3
2 0.5 1.1 2.8
3 0.9 0.02 0.5
4 1.3 0.01 0.06
5 1.7 Not tested <0.01
It can be observed that in both sites, the Cu concentrations in the eluates reduced
significantly after two steps, see Table 11 and Figure 12. Moreover, the same pattern
was detected by both HACH and ICP-AES analyses.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
33 |
Chalmers University of Technology | Table 12 Concentrations of metals (mg/kg) analysed by ICP-AES before enhanced acidic washing (original),
after it (leached) and after washing part (wash 1.7). The weight loss (%) was compensated and measurement
insecurities are not included. The results of the ICP-AES analyses are evaluated against the KM and MKM
guidelines represent on the right. Concentrations that are higher than the MKM guidelines are highlighted in
dark orange.
Substance I II KM MKM
I I II II
mg/kg washed washed mg/kg mg/kg
original leached1 original leached2
DS 1.73 1.74 DS DS
As 0.4 1.2 0.9 1.3 2.1 2.1 10 25
Ba 526 559 547 670 699 726 200 300
Cd 0.0 0.1 0.0 0.1 0.2 0.1 1 15
Co 2.1 2.7 2.1 1.8 1.4 1.3 15 35
Cr 25.7 32.1 33.2 25.7 14.9 19.5 80 150
Cu 2390 650 463 4760 673 446 80 200
Hg <0.04 4.1 2.2 0.0 4.9 6.0 0 3
Ni 3.9 4.8 3.6 3.4 2.7 2.5 40 120
Pb 8.2 9.2 5.4 9.4 7.2 5.6 50 400
V 46.1 51.1 41.3 29.0 23.9 20.3 100 200
Zn 13.0 34.4 14.2 23.4 30.4 15.2 250 5001
I leached - Solid residue left after batch leaching of soil sample from Site I (acidic leaching)
II leached - Solid residue left after batch leaching of soil sample from Site II (acidic leaching)
I washed 1.7 - Solid residue from Site I left after acidic leaching followed by subsequent washing steps (5) with
Milli-Q water (washing), final L/S-value is 1.7
II washed 1.7 - Solid residue from Site II left after acidic leaching followed by subsequent washing steps (5) with
Milli-Q water (washing part), final L/S-value is 1.7
The concentration of Cu was reduced in the final residues around 5 times from both
sites. Nevertheless, in all studied residues the Cu content is still higher than the MKM
guideline. This estimated with the account of soil sample weight loss and
measurement insecurity of ±21%.
The concentration of Hg sharply increased after the acidic leaching for both sites. It is
likely that Hg was adsorbed from the process water. For Site I, its concentration has
decreased from being higher than the MKM guideline to just slightly below it after the
washing part. However, the weight reduction and measurement insecurity of ±30%,
reported by external laboratory, means that Hg concentration in the final residues is
equally likely to be below or above the MKM guideline. Together with soils’
heterogeneity it is hard to make a fair conclusion about the effectiveness of the
washing part in reducing of Hg concentration.
The soils from both sites showed no indicative changes in Ba concentration during the
whole process of enhanced washing with its concentration measured higher than the
MKM. It can be concluded that neither process water nor Milli-Q water is effective in
promoting Ba dissociation. According to IPCS (1990), Ba exists in nature as the
divalent cation Ba2+ and readily reacts with surrounding elements and compounds to
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
35 |
Chalmers University of Technology | form different composites. The only compound formed with Ba, with solubility that
does not increase with decreasing pH is barium sulfate (Ba SO ). Based on this, it can
4
be concluded that both sites had Ba in form of barium sulfate. This also means that Ba
is in a stable form, which is a positive quality when soil residues were evaluated
further.
The concentrations of most substances were lower in residues after the washing step
then those in residues after the acidic leaching, with some remained unchanged. The
washing part was noticeably successful in reducing Cu (1.5 times) and Zn (>2 times)
concentrations.
According to Dermont, Bergeron and Mercier (2007) the removal efficiency depends
on what metal type is to be roved as well as the valence of the element. In both sites
the enhanced soil washing showed a reduction in concentration of Cu, Pb and Zn.
These metals express cationic qualities, thus their dissolution increases when the
solution pH decreases. The same metals showed a good responds to the washing part,
by reducing their concentrations further. It can be hypothesized that they were weakly
associated due to affect of acidic leaching or in a free form trapped in pores.
On the contrary, the solubility of oxyanions of the metalloid such as As decreases
when the solution’s pH decreases (Dermont, Bergeron and Mercier, 2007). This may
explain why the developed enhanced washing method was ineffective in reducing As
concentration. Moreover, there is an increase in As concentration in both sites due to a
possible adsorption from process water enhanced by low pH (Mulligan, Yong and
Gibbs, 2000). Nevertheless, the concentrations of As in all studied residues remained
well below the KM guideline.
Soil matrix weight change
To be able to determine the efficiency of the developed enhanced soil washing
method, the weight change had to be estimated. As a result, it was found that the total
weight loss after the soil was subjected to the soil washing was 11.8 w% for Site I and
7.7 w% for Site II. Additionally, it was assumed that 0.5 w-% was lost not due to
process of dissociation but left behind on the laboratory equipment while samples
were moved and divided. Consequently, the soil weight losses that were taken for
further calculations are as follows:
Site I – 11.3 w%
Site II – 7.2 w%
The degree of weight loss of the soil depends on many factors, similar to those
defining the efficiency of metal extraction from soils, such as soil geochemistry or
extracting reagent used. The calculated weight losses comply with the findings in
study by Karlfeldt Fedje et al. (2013), where by using acidic leaching agents the
weight loss of 10-15 w% was estimated.
Efficiency
To give an ample account of the effectiveness of the developed method, two
efficiencies were calculated:
1 - The soil cleaning efficiency, reflecting the effectiveness of the method to reduce
Cu concentration in the soil (! ):
!
Site I - 81%
Site II – 90%
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
36 |
Chalmers University of Technology | 2 - The leaching efficiency (! ) that reflects the ability of the method to receive Cu-
!
rich eluates:
Site I – 69%
Site II – 86%
The method achieved a high soil cleaning efficiency in both sites, with higher ! for
!
Site II. The calculated leaching efficiency is showed to be higher for Site II also.
Therefore, it can be concluded that the enhanced soil washing method is more
effective in leaching Cu for soil from Site II. Even though the 86% ! is reasonably
!
high efficiency it is still lower than the results achieved in study by Karlfeldt Fedje et
al. (2013). In the latter, 90 - 100% of Cu was released from some soil samples when
process water was used.
Dermont, Bergeron and Mercier (2007) concluded that one of the important factors
affecting metal removal efficiency by the chemical extraction is soil texture and
metals affinity to the soil matrix. Such as one of the possible limiting factors in this
study is the high content of clay/silt. From Table 14 it is clear that soils from Site I
have higher content of clay (6%) and silt (45.5%) then soils from Site II (3% and 9%
correspondingly). Moreover, as stated by Mulligan, Yong and Gibbs (2000) the soil
washing technique is most efficient with sandy soils. These statements has been
confirmed by the results in this study that showed that fine sand coarse silt soil from
Site I have a higher adsorption capacity for Cu contaminant than sand soil from Site
II.
Nevertheless, according to Dermont, Bergeron and Mercier (2008) one of the factors
that may reduce the effectiveness of the chemical leaching is high humic content.
Based on the results in Table 14, it can be seen that the humic content in the original
soils from Site II is twice as high as that in Site I. Therefore, its not always possible to
clearly explain metal removal efficiency since it depends on several aspects and their
combinations.
pH
In Table 13, the changes in pH can be observed after the soil was treated with process
water followed by Milli-Q water.
Table 13 Comparison of the pH of original soils and final residues from Sites I and Site II.
Site I (Köpmannebro) Site II (Björkhult)
Original soil Final residues Original soil Final residues
4.8 4.3 5.0 3.6
It is evident that soils in both sites became more acidic after the enhanced soil
washing, with Site II showing a greater decrease in pH. It can be hypothesised that
soil from Site I had a higher buffering capacity then soil from Site II. According to
Zoltán (2010) the acid-base buffering capacity of organic-rich materials is usually
high. The soils from Site II have humus content twice as high as soils from Site I
(Table 14), which contradicts the result in Table 13. On the contrary, the same source
states that clay soils have higher buffering capacity than sandy soils; in this study soil
from Site I has higher clay content than soil from Site II, see Table 14 (Zoltán, 2010).
Once more, it is challenging to give a certain explanation for one soil being more
sensitive to acid introduction due to soils’ complex chemical composition and
multiple factors affecting its characteristics.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
37 |
Chalmers University of Technology | silt in Site II. The names given to soils according to the SLU (2014) classification
system and reflect the changes that took place after the remediation, see Table 14.
Soil as Filter and Buffer for Heavy Metals
The soil ability to immobilize or neutralize substances by mechanism of physico-
chemical adsorption and via metabolism processes was evaluated using the TUSEC
manual. The results presented in Table 15.
Table 15 Comparison of scores according to the TUSEC method estimated for pre-and post-remedial soils for
soil function ‘Soil as filter and buffer for heavy metals’.
Site I original Site I residues Site II original Site II residues
5 5 4 5
According to TUSEC manual the soil from Site I showed ‘very low’ constitution with
no quality changes between original samples and the final residues. This means that
although the original soil was of a very low ability to buffer and bind/sorb heavy
metals, this did not worsen due to the affect of the developed enhanced soil washing.
For Site II the soil’s capacity to buffer and filter heavy metals decreased from ‘low’ to
‘very low’ after the enhanced soil washing. The main reason is soil residues became
more acidic with pH decreasing from 5.0 to 3.6 (Table 13). Because the binding
ability of clay particles and organic matter is pH-dependent and decreases with
decreasing pH-value (Baird and Cann, 2008), the significant reduction in pH in Site II
led to lower adsorption and binding abilities.
SEM
The metals speciation, using SEM, allowed identifying the changes in major elements
such as those forming the mineral inorganic part of the soil and minor elements such
as Cu. The results of the findings presented in Table 16.
Table 16 Summary showing an average w-% change in some of the soils’ common elements, based on spot
analysis, SEM.
Sample Al w-% Si w-% Ca w-% Fe w-%
I orig 5.6 23.2 0.9 3.6
I resid 4.7 21.9 0.8 12.1
II orig 5.4 16.6 0.3 8.6
II resid 5.2 15.7 0.3 4.1
Site I showed to have a greater w-% loss in Al, Si and Ca elements after acidic
washing. It can be hypothesised that the process water had a stronger impact on the
soil matrix in Site I as it was confirmed by its greater weight loss compared to the Site
II.
In attempts to explain the sharp rise in Hg content in soil residues for both sites the
soils’ mineralogy was assessed. In focus were the major soil constituent elements such
as Al, Si, Ca and Fe. The reasonably high content of the Al and Fe (Table 16, the data
highlighted in grey) found by SEM signifies that soil matrix composed of clay
minerals and iron oxides (Fernández, 2013). Consequently, Fernández (2013) claims
that Fe, Mn, Hg and As tend to associate geochemically. Therefore, it can be
hypothesized that high content of iron oxides in original samples from both sites had a
high affinity towards Hg, which lead to it adsorption from the process water. This
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
39 |
Chalmers University of Technology | conclusion is complementary to Baird and Cann (2008) that stated that Hg, upon
contact with soil, becomes captured because it forms insoluble compound with sulfur
ions that present in clay particles and organic matter.
Additional information was gained by analysing the magnified images, Figure 13.
a) b)
d)
c)
e) f)
Figure 13 SEM images of soil samples a) Original soil, Site I, 200x b) Final
residues, Site I, 200x, c) Original soil, Site I, 2000x d) Final residues, Site I, 2000x
e) Original soil, Site II, 2000x f) Final residues, Site II, 2000x.
The SEM images for both sites show that the post-remedial samples looks more
compact with less smaller individual particles of irregular shapes.
For Site I, there is more particle size variations can be observed in the original sample
while the whole crystalline structure of the soil residues looks more similar across the
sample. The larger agglomerates in the original samples can be an association of
smaller particles adhered together that dissociated after the acidic leaching. Similar
physical changes can be observed in the Site II. The residue sample looks more
porous compared to the original sample. The changes in the orientation of small
fragments could be observed in post-treated sample, which is a possible effect of
sorption mechanisms that occurred during metal dissociation Dermont, Bergeron and
Mercier, 2007.
SS-EN12457-3 leaching test
As the studied soil residues did not comply with the KM/MKM guidelines, the
downscaled standard leaching test was applied – waste characterisation. It is
important to note that not all of the listed substances in the SEPA’s report were
analysed in this project. The recommendations with regards to disposal of the waste
were based only on the substances presented in Table 16.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
40 |
Chalmers University of Technology | Table 16 Concentrations of metals in the eluates produced after the leaching test (SS-EN12457-3) and analysed
by ICP-AES, mg/kg. No highlight = accepted as inert waste; Green = within non-hazardous waste limit; Pale
orange = within hazardous waste limit; Dark orange = above hazardous waste limit.
Non- Above
Inert
Element hazardous Hazardous hazardous
Site I Site I Site II Site II waste
mg/kg waste waste waste
original residues original residues mg/kg
DS mg/kg mg/kg DS mg/kg
DS
DS DS
As 0.6 0.6 0.6 0.6 0.5 2 25 >25
Ba 11 4 16 18 20 100 300 >300
Cd 0.1 0.1 0.1 0.1 0.04 1 5 >5
Cr tot 0.1 0.1 0.1 0.1 0.5 10 70 >70
Cu 123 19 128 23 2 50 100 >100
Hg - - - - 0.01 0.2 2 >2
Mo 0.1 0.1 0.1 0.1 0.5 10 30 >30
Ni 0.2 0.2 0.2 0.2 0.4 10 40 >40
Pb 0.6 0.8 0.6 1.1 0.5 10 50 >50
Sb - - - - 0.1 0.7 5 >5
Se - - - - 0.1 0.5 7 >7
Zn 6 9 9 10 4 50 200 >200
For the analysed metal contaminants, the soil residues from both sites fulfil the
requirements to be landfilled as non-hazardous waste, while the original soils do not
even fulfil depositing in a landfill for hazardous waste. The biggest success of the
developed enhanced soil washing method can be seen from observing the changes in
Cu. In both sites, when the standard leaching test was applied on the residues sample,
the Cu leaching reduced around 6 times. The reduced mobility for Cu, Ba and Zn was
noted earlier, in the washing part optimisation of the batch experiment. Hence, the
decrease in mobility of most metals as showed by SS-EN12457-3 test, can be
explained by combination of reduction in total metal concentration and the stabilizing
effect of the two washing steps with total L/S=0.6.
A slight mobility increase in Zn can be seen for soil residues in both sites, however,
the leached concentration is still well below the non-hazardous limit. The same
pattern of behaviour for Zn was observed by Eriksson and Johansson (2013). There is
an increase in Pb mobility in the final residues compared to the original soils from
both sites. Nevertheless, the concentrations of Pb remain within the same class of
waste for pre-and post-remedial soils in Site I and II.
Based on the downscaled SS-EN12457-3 test and the analysed substances in Table 16,
it can be concluded that soil residues from both sites fulfil the requirements to be
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
41 |
Chalmers University of Technology | disposed as non-hazardous waste i.e. the proposed soil washing method clearly ends
up in cleaner soil residues compared to the original ones.
6.5 Prospect
To get an idea about the prospect of implementing this remedial method on industrial-
scale, the total amount of acidic wastewater produced in Sweden was estimated. The
yearly production of acidic wastewater in Sweden is accounts for 74 million tonnes
(Carlsson, 2010). Consequently, it can be estimated what amount of Cu contaminated
soils can be remediated using acidic process water in a year, see Table 17.
For estimation L/S=8 and dilution ratio of process water of 75/25 were used. For
Köpmannebro (Site I), only the following taken into account: peat (5 tonnes), bark (26
tonnes) and clay soil (2 tonnes). For Björkhult (Site II) the estimation was done for
21900m3 of contaminated soil.
Table 17 Estimation of supply and demand for acidic wastewater in Sweden, tonnes/year.
Production of acidic Demand in acidic wastewater, tonnes
wastewater, million
tonnes/year Site I Site II
74 180 21353
Soil can be washer,
694000
tonnes/year
Although, the above calculations are approximates, they show that the yearly
availability of process water can only satisfy a fraction of the possible total demand of
the 80 000 potentially contaminated sites in Sweden. However, because it is certain
that not all contaminated sites would require the same type of treatment i.e. acidic
washing, less process water would be required per year. Moreover, with the potential
to re-use waste product (process water) and to recover Cu (from eluates) at the same
time, this method may have a positive prospect in the future.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
42 |
Chalmers University of Technology | 7 Conclusions
Based on the results from the laboratory experiments performed while developing the
enhanced soil washing method it can be concluded that soils contaminated with Cu
can be treated with acidic process water to significantly (~90%) reduce Cu
concentration. By doing so, it is possible to produce Cu-rich eluates with the potential
to recover the valuable metal. Also, the use of the method allows producing final
residues that are stable enough to be deposited in a landfill for non-hazardous waste,
which was not possible with the original soil. Together with an associated reduction in
the total volume of the post-treated soil, the use of this technology expected to lead to
lower financial burden linked with waste management.
Additionally, the following disclosures were made:
• Despite that the final residues had Cu concentration 5 times lower than that in
the original soils, they still were not able to comply with the MKM guideline.
• Diluted process water (75/25) was able to leach about the same concentration
of Cu as the non-diluted one.
• Hg contained in process water showed an ability to adsorb onto the soil. To
avoid this effect it is necessary to include an additional pre-treatment for the
acidic process water prior using it as a leachant. Further study of Hg
adsorption mechanisms is therefore advised.
• The enhanced soil washing may cause lowering of pH, which can negatively
affect the soil’s function as filter and buffer for heavy metals, as described in
TUSEC manual. However, in this study, the decrease in buffer ability didn’t
promote further leaching of metals from soil matrix based on the standard
leaching test.
• Washing with water, facilitates further removal of some metal contaminants,
while for other metals it showed not to have any significant effect.
• ‘Clayish fine sand/coarse silt’ and ‘slightly clayish sand’ types of soil respond
similarly to the developed enhanced washing method, with the latter leaching
more Cu when physical pre-treatment is used.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
43 |
Chalmers University of Technology | 8 Continuation of the work
Additionally, several recommendations can be made for continued study in this field
of research:
• Further optimisation of L/S, between 6 and 8, for the acidic leaching should be
considered. It is of a great benefit if reduced amount of leachant can yield a
higher concentration of Cu.
• To strengthen the findings in this study it is recommended to repeat the
washing part (batch experiment) with only one step and L/S=0.6.
• Because the presence of organic compounds influences metal mobility, the
effect of organic pollutants on metal separation must be studied. It is possible
that addressing this factor may lead to greater metal leaching results.
• For further experiments involving soil washing with process water, the
dilution ratio of the latter should be chosen specifically for each treated site.
• The full-scale standard leaching SS-EN-12457-3 test should be repeated with
the analysis of organic pollutants included.
• It is recommended to investigate Ba pollutant and opportunity to decrease its
concentration simultaneously with Cu.
• To sum up, further work should attempt to improve Cu leaching from soil with
a view to achieve the compliance with the MKM guideline without negative
effect of Hg adsorption.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:07
44 |
Chalmers University of Technology | Raw Material Detection System for Rock Crushers
GUSTAV KOLLEBY
JACOB LARSSON
Department of Industrial and Materials Science
Chalmers University of Technology
Abstract
Equipment utilization is an important aspect of any crushing operation. Not only is
this an important factor for increasing production yield and profit, but also to reduce
the resources consumed when operating the equipment. In particular, the thesis aims
to improve utilization of primary jaw crushers used in the aggregate industry by means
of state-of-the-art technologies and methods. Presented in the thesis are novel concepts
aimed to reduce down-time and increase yield of the crusher while creating customer value
throughout the crushing process.
From competitor studies, on-site studies, concept development and evaluation, an idea is
taken to the prototyping stage and tested in a real-world environment at a Swedish aggre-
gate plant. The thesis covers the development of both hardware aspects of the proposed
concepts as well as software methods to achieve improvements of the primary crushing
process.
By leveraging today’s inexpensive, robust and high-performance embedded computing
systems and imaging technologies, the prototype showed great potential in gathering data
of the the primary crushing process. In particular, the prototype was able to successfully
identify material flow and material size, two very important aspects of this crushing stage,
with an accuracy of 100% and 45% respectively.
Keywords: Raw, Material, Detection,System, Rock, Crushers, Embedded, Stereo, Vision,
Segmentation.
v |
Chalmers University of Technology | 1
Introduction
This chapter aims to introduce the thesis problem including its background and scope
of the issue at hand. Furthermore, valuable hypothesis are presented in the problem
formulation which are to be answered in the thesis.
1.1 Background
Roctim AB is a Swedish company that focuses on optimization and control of the rock
crushing process and the rock crushers themselves, in order to increase the production
yield and efficiency. They offer complete solutions for the entire rock crushing process
regarding control and monitoring systems. One area that the company sees a need in, is
the monitoring and detection of raw material fed in to the crusher. As such, there is a
possibility of using new technologies to solve such tasks, which has the potential to reduce
down time of crushers and thus increase their efficiency.
1.2 Aim and Vision of the Project
The aim of this project is to develop a vision based detection system for a primary jaw
crusher. The system will be able to identify if there is a rock causing a jam of the crusher
and then send an alert signal that there is an issue with the production. It will be able to
identify individual rocks, measure the flow rate and estimate their size. This will then be
used for statistical purposes and to predict if a rock would cause a jam, even before it has
entered the crushing chamber. Also, the project will look in to the economic benefits of
the system and what customer values can be met. Furthermore, as Roctim offers control
systems for rock crushers, this solution will be integrated to their existing product.
1.3 Limitations
The project and the development of a prototype will be specifically limited to jaw crushers
used as primary crushers. Additionally, the development of the prototype and gathering
of data for development will also be limited to one field study location. Furthermore, the
market analysis and cost calculation will be limited to the Swedish aggregate industry as
this is the main market for the company. The prototype is not limited to a strict budget
and will be decided during the course of the project, with respect to the potential of the
concept. The technology used should be readily available for the company in order to
reduce risk of compatibility issues. System safety and security of the product, such as CE
markings and data security, is not considered in the scope of this project.
1 |
Chalmers University of Technology | 1. Introduction
Figure 1.1: Cross Section View of Metso Nordberg C-Series Jaw Crusher (Metso, 2018)
1.4 Uncertainties and Risks
In any development endeavour, there is always risks and sometimes underestimated conse-
quences of an action. For this project, a number of risks and uncertainties can be foreseen.
To start, one major risk factor would be that the time estimation for the software de-
velopment might be inadequate. A consequence would be that the software is delayed,
which could result in an unfinished prototype. To lower that risk, a rigorous period of
training will be conducted in the early phase to make sure that the team members are
up to date with the development methodologies. The hardware-software integration is
another potential risk area. Perhaps, some commercially available hardware solutions can
present unforeseen issues or requirements of the system that the team is unable to obtain
or realize. This would result in a waste of resources, specifically time, causing the team
to rethink a part of the solution. A solution would be to test early in the process to make
sure that everything is compatible. This ensures that there is enough time to solve the
issue and quickly move on. A lack of knowledge and skill in the team might pose a risk.
For example, if the team, with the current knowledge, cannot find a suitable solution
for the given task. The consequence would be that the product would most likely not
meet customer demands. Thus, research in to other similar problems or areas would be of
great value to lower this risk. This would ensure that a suitable and tested solution has
a greater potential to solve the problem. The solution would be accomplished by early
research and training to close knowledge and skill gaps. The major risks that could be
identified at the start of the project have a suitable method for avoiding and/or solving
the issue. As such, the project itself could be considered safe but of course unforeseen
risks can pose a problem during the project.
1.5 Scope
The scope of this project is to develop a prototype vision based system for jaw crusher.
The team will, during Q1 and Q2 of 2018, develop the prototype system at Roctims office
using the work break-down structure, which is explained further in chapter 2: Method-
2 |
Chalmers University of Technology | 2
Methodologies
In this chapter, the project work flow is described along with the intended development
methodstobeused. Themethodsdescribedhereareonlyresearchmethodsaimedtowards
generating concepts, as the more detailed methods of software or hardware development
are not yet established. This is due to the potential change of the methods used during
the course of development.
2.1 Work Break-Down Structure
To get a better understand the work flow of the project and to further ease the planning,
the project divided into eleven steps. Each step is elaborated further in more detail. This
includes the work carried out and what methods will be used.
1. Customer Needs
2. Interview customers
3. Competitor and Patent Analysis
4. Litterateur and Technology Study
5. Function Break-Down Structure
6. Idea and Concept Generation for Hardware
7. Concept Evaluation and Stage-Gate for Hardware
8. Software Method Study for Chosen Hardware
9. Software and System Development
10. Prototyping
11. Testing and Optimization
12. On-site Testing
Customer Needs
In order to identify and map the customer needs, site visits will be conducted along side
reviewing of articles and news in trade magazines. This would be an effective way of
obtaining the customer needs, as many aggregate mining sites vary from each other in
terms of layout, way of working and the type of equipment used.
Interviewed customers
During the visits, a number of unstructured interviews were carried out. The reason
for choosing unstructured interviews were to ensure that the interview could feel that
he and/or she could speak freely and feel more comfortable. This led towards more of a
naturaldiscussionduringtheinterview,ratherthangoingthroughalistofstrictquestions.
5 |
Chalmers University of Technology | 2. Methodologies
This meant the valuable information regarding their opinions could first be highlighted
that then could be the basis that then led to some follow up questions.
The risk with this method is that some questions and/or topics can be left out but due
to the surrounding environment but also the fact that the operators could not dedicated
a fixed amount of time as they had to tend to alarms, the decisions was to proceed with
the unstructured interviews.
Competitor and Patent Analysis
To gain insight into the already existing solutions and ideas, some analysis tools will be
applied. These include a patent search, to ensure that the development would lead to an
idea that does not infringe upon other companies patents. Also, a competitor analysis will
be conducted in order to identify possible differentiation points or weaknesses that could
lead to a competitive advantage for our company. Additionally, it would give insight in
to the current situation and provide valuable information regarding technologies.
Litterateur and Technology Study
To get a deeper understanding about the designated area and the available technology
a literature study will be carried out. This include gathering information from articles,
papers, journals and such mediums. Furthermore, the TRL-level investigation needs to
be conducted for the technology in order to ensure that the a chosen technology would
be a viable solution.
Function Break-Down Structure
From the aforementioned steps a function tree can be generated. It is an important step
in the process, as it serves to identify the building blocks of the product at a higher
abstraction level. The main and sub functions are generated such that they aid in the
next step of concept generation. Preferably, functions are generated with modularity in
mind, meaning that main functions are not intertwined but still have the potential to
both be integrated or modular.
Idea and Concept Generation for Hardware
Once the team has a common picture of the problem, the idea generation for the hardware
can begin. This means that all ideas should be considered and evaluated accordingly.
This is preferably done with methods such as brainstorming.(Pahl et al., 2007; Ulrich and
Eppinger, 2012). Followed by this will naturally be the concept generation. This means
that the ideas will be combined into concepts that later will be evaluated.
Concept Evaluation and Stage-Gate for Hardware
When a suitable number of diverse concepts have been generated, with respect to the
hardware, they are ranked based on how well they fulfill the customer needs. First,
6 |
Chalmers University of Technology | 2. Methodologies
conceptsarequicklyscreenedthroughaneliminationmatrixinordertoreducethenumber
of concepts into a more manageable amount, using more basic evaluation criteria(Pahl
etal.,2007,Chapter3). Then,theyarefedthroughaModifiedPughMatrixandcompared
with quantifiable customer needs(Ullmann, 2010, Chapter 8). The subsequent results are
then taken in to account and one final concept is chosen.
Software Method Study for Chosen Hardware
Since the hardware is to be locked down, research in to software and methods is required.
This includes the general method that the hardware will require and the information will
be gathered through articles, journals, sample codes, online courses and such.
Software and System Development
When familiar with the concepts for each function, a more detailed software architecture
is developed in order to get an overview of how the different modules will interact. Fur-
thermore, the software modules are worked on in more detail and with more functionality
in mind as well as overall system compatibility.
Prototyping
Once the initial development of the software system is done, the finalized hardware and
the software can be combined. Now the prototype can now be assembled and checked for
potential unforeseen compatibility problems.
Testing and Optimization
At this stage the first testing of the complete prototype can be initiated. The first phase
of the testing will be conducted in a controlled environment were the whole system can
be analyzed immediately and the influencing factors can be controlled. Concurrently will
the optimization of the system be accomplished in order to ensure that the system will
ready for the on-site testing phase.
On-Site Testing
Once the prototype has been optimized with the estimated parameters from the field, the
prototype will be place on rock crusher on-site. This means that prototype will be tested
in a operational environment in order to verify the design and the functional performance
level. The information gathered at this stage will be the basis for the evaluation of
the prototype as well as how the prototype can be even further optimized for future
development.
7 |
Chalmers University of Technology | 3
Literature study
The aggregate industry has been and still is one of the most basic and essential part of
today’s society. The extraction of aggregate products has been improved over centuries
and the products is used in almost everything around us today, everything from roads to
buildings. Between 2015 to 2016 the Swedish aggregate industry increased their total de-
livery by 2 million tonnes, landing on a yearly total delivery of 86 million tonnes in 2016.
The need keeps increasing as the aggregate products is one of the most extracted raw
material in Sweden. The aggregate industry in Sweden has benefited from an increased
demand for crushed rock and aggregates, mainly from the construction sector. Accord-
ing to SGU (Sveriges Geologiska Underso¨kning) and Boverket, the yearly construction
rate for buildings is estimated between 50-55 000 per year. In terms of aggregates, that
would require a yearly production of 100 Mton per year. Looking further in to the future,
Boverket forecasts an increase of construction of 70 000 new buildings per year up until
2025. That would amount to an aggregate production of 120 Mton per year. This means
that there is a need to increase the production in order to avoid the need to import of
the aggregate products. (Norlin and Go¨ransson, 2018) This is not only important from
economical point of view but also from an environmental perspective. Environmental
regulations are of concern both for aggregate plants and for the transportation of aggre-
gates. As such, plants would benefit from better utilization of resources, energy efficient
processes and being closer to construction areas in order to minimize the transportation
emissions. As an example, transporting aggregates by truck for 15-30 km would amount
the same emissions as it took to crush the material at the plant. Due to these aspects, ag-
gregateproducersarelookingtoincreasetheequipmentavailability, equipmentutilization
and become more energy efficient. Same goes with the cost, at a transportation distance
of 30-50 km would the transportation be equivalent to the actual material cost.(Kristian
Schoning, 2017) Another aspect to consider when looking into the aggregate industry is
the transformation that has been going on over the last decades. Since 1985 to 2016
crushed bedrock has been increasing from only 20% of the total material, up till today
where it stands for over 80% of the total material used. The reason for this is that before
sand and gravel was the main product that was used, which is a limited resource that have
a much higher value than crushed rocks. This can also be seen by looking at the number
of quarry permit, that has gone from over 4000 to less than 500 in the last decades. How-
ever, the number of quarry permit for the crushing industry has not increased as much,
only a couple of hundred. Meaning that they need to produce much more and be more
efficient to hold up the production that is needed. (Norlin and Go¨ransson, 2018)
The general process for the aggregated industry is as follow and can be seen in fig. 3.1
were some of the key components of a crushing process is shown. (1) The upstream stages,
consists of both transportation of the blasted rocks and the blasting itself. (2) The feeder
and primary crusher, consists of feeding blasted material in to the crusher and reducing
it’s size by crushing. After this the material either run through a screen to sort out certain
9 |
Chalmers University of Technology | 3. Literature study
size and/or to be placed in a stock pile. (3) The secondary crusher, in order to reduce
the rocks into finalized products it usually go through a number of crushers depending on
the setup, these have a higher accuracy but are more sensitive to incoming rock size. (4)
Unloading area, customers comes to buy different product sizes, depending on the usage
area.
Figure 3.1: Overview of the a crushing process. 1: Blasting and transportation. 2:
Primary crushing. 3: Secondary crusher. 4: Outgoing material
This process is highly dependent on many factors and may vary from site to site. First
the rock gets blasted into manageable pieces, the size here does vary a lot and maximum
allowed size is heavy depending on what the primary crusher are specified for. If the
rocks are to big, they will cause jamming and if they are to small they will not be crushed
in the primary crusher, leaving the crusher as an unused source. After the primary
crusher the rocks has decreased in size were the size is more known but still may vary
depending on the primary crusher and how the rock falls through. After this step, there
are many options regarding processes that can be used. Usually, there are two or more
crushers that will bring down the size of the material that is more commonly used in the
industry. Furthermore, sieves are used between different steps in the process to better
ensure what size the rock has before it enters the secondary and/or the tertiary crusher.
There are different prices for different sizes regarding the end product, which mean that
the producer wants to ensure that the same size it kept even though the crushers wears
out after a certain point. This is where the industry has improved once it comes to using
technology, more specific control systems for the crushers.
As for today, the industry is using many types of different sensors to ensure that the
production can continue to run and ensure that as few stops are achieved, but of course
alsotooptimizetheproduction. Manyofsensorstodayaresuchaslevelsensors,vibrations
sensors, temperature sensors. All to ensure that, for example the secondary crusher runs
as optimal as possible. Also, other sensors such as metal detectors are used to prevent
stops and unnecessary ware and/or damaged on the machines. However, even though
10 |
Chalmers University of Technology | 4
Customer needs study
A dialog with potential customers is an important step towards generating new and cus-
tomer focused products in today’s market. This chapter presents the customer needs
mapping that underlies the product development phase of the thesis.
4.1 On site study: NCC Tagene
The data gathering for the thesis was conducted on site at the local aggregate producer
NCC Tagene, picture of quarry can be seen in fig. 4.1. The site is located 5 km north east
of Gothenburg and produces aggregates and asphalt for local construction projects. They
blast and crush material from their quarry and take care of asphalt recycling as well. The
site consists of one jaw crusher for primary crushing application and cone crushers for the
second and third stage. The primary crusher is located at the highest level of the quarry.
The setup consists of a jaw crusher, a grizzly feeder, and a hydraulic rock breaker as
shown in fig. 4.2. The single haul truck, which runs up and down the quarry, feeds blast
rock directly on to the grizzly feeder. The crusher and feeder are encapsulated inside a
building, which also contains the control room for the whole site.
Figure 4.1: The NCC Tagene quarry. Figure 4.2: Primary crusher with feeder
and rock breaker.
4.1.1 Interviews
During the interviews, a considerably amount of valuable information was gathered. Not
only about what the workers thought was the problems today but also the general way
of working and the general attitude towards certain tasks. The result of the interviews
13 |
Chalmers University of Technology | 4. Customer needs study
showed that, as in many other industries, the workers were getting use to problems such
as jamming and often refereed to that “It is my job to fix it” while other areas, that they
didn’t have control over, were more highlighted. The most relevant information gathered
can be found in the point-list below.
• According to the workers jamming was not a big problem since “it only take around
5 minutes to clear it”.
• They claimed that the flow of the incoming material were often inconsistent and by
that causing the primary crusher to be idle, mainly due to the fact it would run out
of material.
• The size of the rocks is very dependent on the result of the blasting but also when
the material is taken from the blasted section. The first material taken after the
blasting, usually had larger rocks and caused jamming. While material taken in the
end of the blast section, the size of the material shrunk and could go right thought
the primary crusher without the need to be crushed.
• Thecurrentsystemthatareinstalledtokepttrackofthelevelwithinthecrusherhad
some problems, as previously mentioned, the senors was from time to time covered
with dust/dirt and therefor indicated that the crusher was full, even though it could
be empty.
4.1.2 Observations
Throughout the course of the thesis several visits to NCC Tagene have been made. The
purpose was to inspect the primary crusher over a period of time in order to obtain
untold information about the process. Several problem areas have been observed and
documented, such as material flow, size distribution, jamming, equipment downtime and
general issues with the process, which are listed below.
• Some sort of jamming of the crusher seems common, for example too large rocks
wedging themselves over the crushing chamber blocking other rocks to fall down.
These common issues are usually solved within 15-30 minutes. As the operators
mentioned in the interviews, jamming usually takes at most five minutes to clear
with the rock breaker. However, to detect a jam takes longer time. As the operators
are often focusing on other equipment malfunctions or alarms, the crusher might be
jammed for a longer period of time. As such, equipment downtime of around half
an hour may be more common than the operators realize.
• Material flow is highly dependent on the blast fragmentation size. On some visits,
there have been many oversize rocks causing either a jam or a decrease in outflow
due to longer crushing time. On other days, there have been a lot of undersized
material being fed in to the crusher. This causes unwanted wear on the plates and
lowers the equipment utilization. Also, this is an indication of the blasting process.
Alotofundersizedmaterialisusuallytheresultofusingtoomuchexplosivematerial
and drilling blasting holes too tightly together, which causes resources to be wasted.
• The control system for the feeder may be poorly adapted to this crusher. When
the crusher is full, the feeder stops and waits for a lower material level. However,
there was a significant start/stop delay, which caused he material outflow to be very
uneven at times.
14 |
Chalmers University of Technology | 4. Customer needs study
4.2 Customer needs analysis
From on site interviews and the observations the primary problem today is that the
primary crusher is underutilized, which means that the crusher is not used to its full
capacity when crushing material or that it is unable to crush material. This is due to
several factors, which are categorized in the cause-and-effect diagram, figure 4.3, below.
Operatorawareness Blasting
Faultawareness Under/overusage ofexplosives
Response time
Under/overusage of drilling
Lack of process knowledge
Underutilization of
primary crusher
Lackofsituational awareness
Inconsistent material dumping
Manual control
Under/oversized material False sensor readings
Material inflow Control system
Figure 4.3: Cause and effect diagram for the primary crusher
The crusher is properly utilized when there is a steady flow of appropriately sized blasted
rock fed in to the crusher. Every situation that causes a production stop or delay accounts
for significant losses in potential revenue. Operator awareness, blasting, differing material
inflow and the control system of the primary crusher are the main contributors to the
machine being underutilized. For example, if an operator is occupied by another machine,
they are usually not aware of other failures at the time. As such, their response time to
the primary crusher may cause significant delays in production.
What seems to be a significant cause for delays in production is the blast rock size vari-
ation. If the rock is oversized, the crushing time is increased and there are risks of the
crusher being jammed. When the material is undersized, the crushing would be a re-
dundant step in the process, as no value-adding work is being done by the machine.
Furthermore, inappropriately sized material causes unnecessary wear on the plates, which
mayreduceswearpartlife. Also, iftheresultoftheblastingismainlyundersizedmaterial,
there has been a significant waste of drilling time and explosives.
In fig. 4.4, some of the key components of a crushing process is shown. The main process
stages are of concern is the upstream stages (1), primary crushing stage (2) and down-
15 |
Chalmers University of Technology | 4. Customer needs study
stream stages (3). Stage 1 consists of both transportation of the blasted rocks and the
blasting itself. Stage 2 consists of feeding blasted material in to the crusher and reduc-
ing the size by crushing. Stage 3 involves crushing the rocks to even finer end-products.
From stage to stage, material size and material flow are the key components of the plant
performance.
Figure 4.4: Overview of the a crushing process. 1: Blast material in-flow. 2: Primary
crushing. 3: Crushedmaterialout-flow. 4: Feeder. 5: Jawcrusher. 6: Hydraulichammer.
7: Stockpile. 8: Secondary cone crusher.
4.3 Customer needs
When developing a product, it is crucial to specify the needs of all involved parties, in
this case the customer NCC (C) and the user/crusher operator (U).Thus, there is a need
to concretize the given problems and factors that are previously mentioned from both the
observation and the interviews. These comes from both outspoken and unspoken needs
and/or problems that have or will have an affect on the crushing process. The needs
are presented in table 4.1 where the needs will then be given value between one and five
depending on the importance of the need, where one is the lowest and five is the highest
importance level. Each need is paired to multiple issuers. As such, it is important to view
each instance from all perspectives.
16 |
Chalmers University of Technology | 5
Technology study
In recent years, non-contact measurement techniques have become an important topic
for sensing different aspects of an environment. For the aggregate industry, non-contact
measurement systems are highly suitable for the type of tasks presented in this thesis, as
they can be designed to be highly adaptable and in some cases more affordable than tra-
ditional solutions. As such, the technology study focuses on vision and 3D technologies,
for example LiDAR and camera systems. The identified technologies that were feasible
for the task are:
• 2D camera technology
• 3D camera technology
• LiDAR
• Structured-light 3D scanner
• Time-of-flight camera
5.1 2D camera technologies
The rapid pace of camera sensor development and the relatively low cost makes a camera
system a powerful and affordable solution for general sensing tasks. A common usage
scenario for single camera setups is object detection, tracking, segmentation and object
classification. The application areas range from detecting pedestrian on roads to medical
applications to study various samples from patients. Furthermore, distance can be mea-
sured by knowing dimensions and distances to certain objects in a scene. If these criteria
are known, one can study that plane of known distance and then calculate the size of
surrounding objects in that plane. However, if objects are not in the same plane the
measurement results are inaccurate (A. Criminisi, 2002). An analysis of this technology,
when applied to the aggregate industry, is presented in fig. 5.1 below.
19 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Readily available hardware solutions • “You get what you see”
• Highly adaptable to different environments • Requires objects of known size to calculate
• Very inexpensive size of surrounding objects
• Inaccurate size measurements
• Able to extract a variety of aspects regarding • Sensitive to rapidly changing light condition
rocks such as colour, relative orientation and • Signal noise under low light conditions
it's silhouette shape depending on software
• Other usage areas around site
Opportunity Threat
Figure 5.1: Analysis of single camera
5.2 3D camera technologies
As the aforementioned viable 2D imaging technologies inherently lack accurate depth
perception one must look at 3D technologies to obtain appropriate information of the
environment. There are many different technologies with varying types of ways to obtain
a 3D image. However, the end result is the same: a point cloud or surface.
5.2.1 Stereo camera system
To study a 3D space with a camera system, two or more cameras can be combined in
to what is called a stereo camera. The addition of one or more cameras enables depth
sensing, through triangulation of points in both images. The result of such a process
is a depth image with pixels mapped to real world x-y-z coordinates, which can then
be translated to a point cloud (R. Szeliski, 2010). An analysis of this technology, when
applied to the aggregate industry, is presented in fig. 5.2 below
20 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Highly adaptable to different environments • "What you see is what you get“
• Inexpensive relative to other 3D sensors • Requires high computational power for
• Good size measurement accuracy dense point clouds
• Depth map quality highly depends on the
algorithm
• Able to extract a variety of aspects regarding • Sensitive to rapidly changing light condition
rocks such as colour, relative orientation, • Signal noise under low light conditions
silhouette shape and depth data depending • Customers may need real time processing
on software • Mayrequire specialized camera system
• Other usage areas around site
Opportunity Threat
Figure 5.2: Analysis of stereo camera
5.2.2 3D LiDAR
LiDAR is a Laser ranging systems and is a common way of sensing the environment in
variousapplications,forexampleself-drivingvehicles. Theseapplicationsrequirereal-time
performance, stable measurements, long range and high accuracy, all of which LiDAR can
perform well in. However, system cost is high but will most likely be reduced significantly
due to the automotive industry fast paced development (Himmelsbach and Wunsche,
2008). Ananalysisofthis technology, whenapplied totheaggregateindustry, ispresented
in fig. 5.3 below
21 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Highly adaptable to different environments • "What you see is what you get“
• Very good size measurement accuracy • Low field of view
• Real-time performance • Very expensive
• Stable depth map
• Extensive use in other fields • May require additional camera if more scene
• Possible to use outdoors and under any light information is needed
conditions
• Other usage areas around site when
combined with a camera
Opportunity Threat
Figure 5.3: Analysis of LiDAR
5.2.3 2D LiDAR
2D line scanners are used widely in the manufacturing industry to provide detailed scans
of parts, in order to ensure product quality. As the system relies on 2D line scanning,
a 3D surface can only be obtained when either the scanner is moved along a surface or
when the object is moving itself. As such, to properly measure objects they have to move
at a constant speed (Himmelsbach and Wunsche, 2008). An analysis of this technology,
when applied to the aggregate industry, is presented in fig. 5.4 below.
22 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Very good size measurement accuracy • "What you see is what you get“
• Real-time performance • Low measurement range
• Stable measurement • Requires continuous scanning to obtain
point cloud
• Requires constant material speed
• Veryexpensive
• Used in the field • May require customized mounting solution
• Possible to use outdoors and under any light for each site
conditions
Opportunity Threat
Figure 5.4: Analysis of 2D line scan
5.2.4 Structured-light 3D scanner
Structured light systems project a known pattern on to a surface and investigating the
distortion of this pattern with a camera. Based on the distortion a 3D map can be com-
puted. However, multiple projection-reading cycles with different projections are usually
required in order to obtain a good depth image. Of course, such a process substantially
increases the scanning time, but results in a very accurate depth map. Furthermore, the
projectionmaybesensitivetointerferencefromthesurroundingenvironment, forexample
in bright conditions where the projection may be “washed out” by ambient light (Kutu-
lakos and Stegere, 2005). An analysis of this technology, when applied to the aggregate
industry, is presented in fig. 5.5 below.
23 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Good size measurement accuracy • "What you see is what you get“
• Stable measurement • New technologyinthefield
• Requires specialized hardware
• Poorreal-timeperformance
• Complexsoftware
• Expensive
• Possible to use under low light conditions • May needtofullyencapsulate thesystemto
• Improvedresultswhenobjectsarestationary reduceinterferenceofprojection
ormovingslowly due to less blurring effect • Real-timeperformancemaybe too slow for
other applications around the site
• May require additional camera if more scene
information is needed
Opportunity Threat
Figure 5.5: Analysis of 3D structured light scanner
5.2.5 Time-of-flight camera
Time-of-flight camera systems work similarly to ultrasonic sensors, but using light pulses
instead of acoustic pulses and a camera sensor instead of an acoustic receiver. The time
pulse timing and the resulting reflection of light will determine the distance to objects,
as closer objects reflect light quicker than distant objects, which translates to a very easy
distance calculation. Although a very simple concept in theory, in practice it requires
high speed sensors as the reflection time of light is very short, which may require more
performance than cheap camera sensors can provide today (L. Li, 2014). An analysis of
this technology, when applied to the aggregate industry, is presented in fig. 5.6 below.
24 |
Chalmers University of Technology | 5. Technology study
Strength Weakness
• Good size measurement accuracy • "What you see is what you get“
• Real-time performance • New technology in the field
• Stable measurement • Requires specialized hardware
• Simplealgorithm to calculate depth • Expensive
• Possible to use outdoors and under many • Possible interference of light pulses
light conditions • May require additional camera if more scene
information is needed
Opportunity Threat
Figure 5.6: Analysis of Time-of-flight camera
5.3 Competitor analysis
In order to ensure that the product can be a viable and compete with other system on the
market,abenchmarkwerecarriedout. Today,thereareaonlyahandfulsolutionsthatcan
achieve the same or similar task. Also, other industries were used for inspiration such as
thefoodindustryusesthesameand/ordifferentsolutionstoinspectthequality. However,
these solutions is highly specialized and often placed in a well controlled environment,
which is the reason they are not considered in the benchmark. The benchmark will be
focus on which tasks it can solve as well as the accuracy of the product. As the benchmark
shows, some products is more focus on solving one or a few tasks that some part of the
aggregateindustryrequests. Alsosomesolutionsareheavilydependentwereintheprocess
it is placed. The price for most of these solutions are unknown and highly dependent on
many factors. One of them being that most of these solutions are relatively new to the
market with new technology which means that price can vary from a few thousands to
over hundred of thousands euros.
5.3.1 ScanMin Africa
The solution that ScanMin Africa offers is a “Oversize Detection System” (ScanMin
Africa,2018)thatisfocusonmonitoringandpreventingblockageinthesecondarycrusher.
Theproductalsohavetocoverthewholeconveyorbeltinordertofunction, henceitneeds
a controlled environment. However, there are not much information from the supplier re-
garding the actual technology or the price. This does not necessary means that the
25 |
Chalmers University of Technology | 5. Technology study
competitor should be left out and something to keep in mind in the further.
5.3.2 MotionMetrics
MotionMetricsisacompanythatoffersfourdifferentproductswereallusesavisionsystem
to solve various task. Two of the solutions detecting missing tooth and the surrounding
area around excavators and loaders. While two of them are more focused on the usage
of measure the size of rocks. One of these two solutions is a handhold portable solution,
PortaMetrics, that the user can take with them and analyze rocks anywhere. However,
the resolution on the cameras is a bit on the lower end which might affect the accuracy on
the system.(Motion Metrics, 2018b) The other product is a fixed solution that is installed
on the belt. According to MotionMetrics, the solution can detect rocks as small as 0.6
cm with and accuracy of ± 10 %. However, this solutions have a long processing time
and with a low capability of only 2 images per minute. This is most likely to low for a
primary crusher due to the uneven size distribution of rocks. (Motion Metrics, 2018a)
The price for Beltmetrics is said to be around 70 000 EUR but is very dependent of the
site. Also the PortaMetrics can be used with an yearly subscription of 12 000 EUR.
5.3.3 MBV-Systems
MBV-Systems offers a systems to 3D-scan the material on the conveyor belt (MBVsys-
tems, 2018). This creates a 3D-map of the material that can be used to analyze the size
and the distribution of the material. This solution only focus on the material on conveyor
belt since it requires a special installation within a somewhat controlled environment.
The system needs to be a fixed installation around one meter above the conveyor belt.
The price of the system remains unknown since it will most likely be highly dependent
on were and how it will be installed.
5.3.4 Split-Engineering
Split-Engineeringoffershandfulofproductsthatfocusonanalyzingthesizeofrocks.(Split
Engineering, 2018) Their system can according to themselves be placed at every step of
the process in order to analyze the size of the rock and furthermore use this information
to validate the different processes. Two of their solutions focus on a more robust camera
that can be place on either the excavator, “Split-Shovelcam”, or on place that it is known
that the trucks stop at, “Split-Truckcam”. The information regarding the camera system
is very limited but every system needs to be calibrated at site. Which is mentioned in
chapter 5.1.1, which most likely mean that they use a single camera to capture the images.
Regarding the information about the price for different system, is as for most companies
earlier discussed, hard to find. However, the price for a licence is about 8000 EUR which
gives an indication of which price region they are placing them self in, this does not
include any camera or installation costs.
5.3.5 Stone Three Mining
Stone Three Mining is a company that offers four solutions that is specialized in different
areas.(Stone Three Mining, 2018) Two of the solutions focuses more on the extraction of
minerals such as flotation sensors and bubble sizer. The other two solutions is however
26 |
Chalmers University of Technology | 6
Hardware development
6.1 Hardware requirements
The requirements is an essential part of the development process. It will ensure that all
involved parties knows how the product will preform as well as ensure that the product
can compete on the market. It is also used to narrow down and evaluate the concepts
in the concept screening. The basis of the requirements comes mainly from the customer
needs and the competitor analysis. It is also important to quantify the data so it can
used for validating the product in order to ensure that the performance level is met. The
full requirements list can be found in appendix A while a shorter version with some of
the highlighted requirements can be found in table 6.1.
This product has its focus towards the primary crusher and the requirements will be set
around that assumption. Since the primary crusher is just one part of the entire process
chain, it would be desirable to include other machines as well. Furthermore, since it is a
harsh environment it comes natural that the vision system needs to withstand this. There
for it is going to need an IP classification of at least IP65. However, the complete system
with the processing power can in a control room or similar but it would be desirable to
have an IP classification of 68 to be able to place it anywhere at the site.
As the benchmark showed most systems were priced around 10 000 EUR and upwards.
However, the systems are highly dependent on unique mounting position which will most
likely increase the price. Since Roctim offers products to both small and large companies
it would be beneficial if the price could be at the same price range as other product.
Therefore the maximum production cost was set to 15000 EUR, this will not only ensure
that price is well below other competitors as well as giving the opportunity to offer this
product to both small and big companies.
Table 6.1: Accuracy target requirements
These requirements are set with the knowledge gained from all steps in the development
process such as customer needs, technology study etc. The remaining requirements R8 to
R12 takes the performance of the system into consideration. It is most important to set
a minimum performance level that needs to be satisfied. Furthermore, it also important
to consider inaccurate measurements. The reason why this is important is to ensure that
29 |
Chalmers University of Technology | 6. Hardware development
the operator can trust the notifications he receives. If the system falsely indicates that
the crusher is jammed too often, the operator will most likely pay less attention to the
system.
6.2 Function analysis
The main idea of thisthesis is to improve the primary crushing process with new technolo-
gies that provide data which aids in the plants strive for continuous process improvement.
The function tree was chosen as a method for analyzing the functions, as it gives a clear
view of what affects the main function and how to achieve the main function’s goal. The
full function tree can be viewed in appendix B. In this chapter, a more detailed descrip-
tion of each function is provided along with the reasoning behind them. As the intended
system would be used as an analysis tool for decision making, the human functions are
listed as blue functions while the analysis functions are marked with orange.
As the primary jaw crusher is the focus, maximizing it’s performance is the main function
of the system. As such, the overarching function to solve is “Maximizing the jaw crusher
performance”. This would give a direct impact on the utilization of the crusher.
Figure 6.1: Function tree: Upstream process
The material being fed into the crusher is a crucial part of both the plants resources
and investments. As such, the first sub-function is “Optimize blast material in-flow”.
In order to improve this process, one can identify two aspects that affect the material
in-flow. Firstly, “Improve blasting”, which is directed towards producing appropriately
sized material for the crusher. If the blasting is too efficient, the blast material size is
very small. Thus, most of the material fed in to the crusher tends to just pass through.
This results in a crusher that does no value-adding work. Furthermore, this also means
that the blasting process used too much explosive material and/or that they drilled too
many holes. All of these aspects translate to real world costs and are a large part of the
actual operating costs of the plant.
To improve the blasting there needs to be concrete data and good communication with
the blasting crew, which are presented with the functions “Inform blasting crew” and
30 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.