Dephosphorisation of Acidic Wastewater

IN

DEGREE PROJECT TECHNOLOGY, FIRST CYCLE, 15 CREDITS

,

STOCKHOLM SWEDEN 2018

Dephosphorisation of Acidic

Wastewater

Aimed to allow the recirculation of byproducts as

slag builders

MARTIN PALM

HENRIK WELANDER

KTH ROYAL INSTITUTE OF TECHNOLOGY

Abstract

This study was conducted to evaluate if Polonite® could remove Phosphorus from Sandvik’s wastewater that naturally forms during steelmaking, and hence be proficiently used as a precursor of the neutralization process of said steel plant.

Currently lime (CaO) is mixed with the wastewater which creates clean water and sediments, that later need to be landfilled. Due to the high Phosphorus content in the sediments these cannot be recycled back into the production, this result in a loss of important elements.

The aim is therefore to remove the Phosphorus to a target level of 1-2 [mg/L]. In fact, the Phosphorus present in the wastewater drifts in the solid particles derived by the neutralization process, impairing their possible recirculation as slag builders. To evaluate the purifying abilities of Polonite® a batch test and a column test were conducted. During the column test measurements of pH and conductivity were made. To analyze the composition, Inductively Coupled Plasma (ICP) tests were performed.

No clear correlation between the pH and the conductivity of the water could be established from the results. Overall the Polonite® removed high amount of Phosphorus from the wastewater. The target level of 1-2 [mg/L] Phosphorus was achieved in the batch test. Although, Polonite® also incidentally absorbed most of the Fluorine present in the

wastewater making it a questionable choice for the process, as Fluorine is desirable in large quantities in the sediments if used as slag builders. As no ICP results were received from the column test only pH and conductivity are discussed.

Sammanfattning

Denna studie utfördes för att utvärdera om Polonite® kunde rena Fosfor från Sandvik’s avfallsvatten som naturligt uppkommer från stålprocesserna, och använda det som ett försteg i neutraliseringsprocessen.

Idag blandas kalk (CaO) med avfallsvattnet vilket resulterar i rent vatten och sediment, dessa deponeras sedan. På grund av den höga halten Fosfor i sedimenten kan dessa inte

återcirkuleras in i produktionen, detta leder till att viktiga ämnen går förlorade.

Målet är därför att sänka halten Fosfor till en nivå på 1-2 [mg/L]. Fosfor i avfallsvattnet driver i de fasta partiklarna från neutraliseringsprocessen, vilket förhindrar möjligheterna att återanvända dem som slaggbildare. För att utvärdera Polonites® reningsförmåga gjordes ett kolumntest och ett skaktest. Under kolumntestet mättes pH och konduktivitet. För att analysera innehållet i vattnet gjordes ICP-tester på det renade vattnet.

Inget tydligt samband mellan konduktiviteten och pH kunde fastställas från resultaten. Överlag hade Polonite® en hög grad av rening av Fosfor. Den givna mängden av Fosfor uppnåddes i skaktestet. Olyckligtvis hade Polonite® även en hög reningsgrad av Fluor som även fanns i avfallsvattnet, vilket gjorde Polonite® tvivelaktigt försteg i denna process, då Fluorrikt vatten är önskat för att kunna användas som slaggbildare. Då inga ICP-resultat har mottagits diskuteras enbart pH och konduktivitet för kolumntesterna.

Abbreviations

CaCO3 Calcium carbonate

CaF2 Fluorite

CaO Lime

H2SO4 Sulfuric acid

HNO3 Nitric acid

Col.1 Column 1 Col.2 Column 2 Col.3 Column 3

A, B, C, D 1-3 Flasks with 5, 10, 15 and 0 [g] Polonite® P1-6 Period 1-6

P6 N Period 6 night time P7-8 Period 7-8

Table of Contents

2.6.3 Mixed acid pickling ... 5

2.7 Neutralization of wastewater ... 5

2.8 Inductively Coupled Plasma ... 5

1

1 Introduction

Formation of wastewater is a natural part of the steelmaking process, as water is used both for cooling and to remove excess pickling liquor [1]. The pickling process is used to remove the utmost layer of steel which can contain high amount of impurities, rust or other

contaminations [1,2].As Sandvik’s steels are custom made for specific applications they contain many different elements such as chromium, titanium, nickel, zirconium and

molybdenum [3]. The combined wastewater from all pickling processes will therefore contain a wide range of elements, in this study the focus is phosphorus (P). In order to release the wastewater into the environment, its acidic content needs to be neutralized. Currently lime (CaO) is mixed with the wastewater, which reacts with the acids present increasing the pH to an acceptable value. The products of the reaction with lime are clean water and solid

particles. These particles, called sediments, are then filtrated from the clean water and put in landfills. Today the sediments from the purified wastewater cannot be reused, mainly because of the high amount of P which even in small amounts can be harmful, making the steel more brittle [4]. If P would successfully be removed from the wastewater the sediments could be reused. The sediments contain many useful compounds that could be brought back and used in the steel production, one of these key elements being fluorine, F.

P has already been successfully removed by using Polonite®, as shown in “Characterization of Opoka as a basis for its use in Wastewater treatment” by Z. Brogowski and G Renman, presenting a maximum sorption of about 119 [g/kg] phosphate, PO43-[2]. The study was

conducted on synthetic P solution, meaning that the sorption should most certainly differ in this experiment. As Sandvik’s wastewater greatly varies, in composition, pH and water flow, it was difficult to determine the amount of Polonite® needed to purify it to low enough levels of P. To minimize the need of adding additional fluorine, in the form of CaF2, F was desired

to remain after the purification of the wastewater.

2

2 Literature Study

2.1 Phosphorus

Phosphorus, P, is a basic element that is part of the nitrogen group in the periodic table. It is a non-metallic element that is not found pure in nature. Phosphate ore is the most common form that P is found as. P is essential to all living things as it forms the backbone of DNA. It is also important in cells for energy transport as adenosine triphosphate (ATP), and many other biologically important molecules [9]. P is often seen as an impurity in the steelmaking process because of its embrittling effects on steel. Although P improves machinability and increases the tensile strength of steel, it can also generate a range of negative effects

depending on the amount added. In higher grades of steel the maximum amount of P allowed is between 0.03% and 0.05% due to its detrimental effects. For low-alloy high-strength steels the maximum rises up to 0.10% as the P strengthens as well as improves its corrosion

resistance. The embrittlement increases when the content of P in hardened steel gets too high. The strength and hardness improves as the ductility and toughness decreases [10].

P is one of four macro-nutrients that plants require to thrive, as they need relatively large amounts for healthy growth. P is one of the essential elements for the process of

photosynthesis, energy transfer and nutrient transport. A plant grows more vigorously, will mature earlier and has a higher productivity, if it has proper amounts of P to nourish from [11]. Today we extract roughly 160 million tons of P per year, where most of it goes into the agricultural industry as artificial fertilizer [12].

2.2 Fluorine

Fluorine, F, is a basic element in the halogen group in the periodic table. F is crucial in steelmaking, as fluorite (CaF2) could be used to decrease the viscosity of slags [13]. This

enables faster transportation of mass and heat and it lowers the melting temperature, while also minimizing the loss of chromium (Cr). CaF2 also stabilizes the slag and increases the

phosphate capacity for slags with high basicity containing CaO. In high basicity slags fluorophosphates (3Ca3(Po4)2CaF2) are formed, which dissolves oxides and lowers surface

tension between slag and the melt [13].

CaF2 has great qualities but is not commonly used, as it poses environmental implications.

The F is extremely reactive, the gas that forms is a deadly poison and combined with hydrogen (H) it can form hydrofluoric acid (HF) [14]. HF is considered one of the strongest inorganic acids and can cause severe burns [15].

2.3 Polonite

3

material, it could be reused as fertilizer and it would be a good substitute to other purifying agents, such as various slags [5]. The high porosity gives the silica-calcite rock the ability to efficiently purify water from P and phosphates. As it contains considerable amounts of CaCO3 and SiO2, it was suggested to be a good sorption of P, but beyond the P, other metals

could react to it as well. This might affect its current application as a fertilizer, since it might contain unwanted elements after its use as a purification material for wastewaters [5].

Depending on the original composition of the opoka, it is called either “light opoka” or “heavy opoka”. The light version contains more SiO2 and is more porous, with higher CaCO3

content it’s referred to as heavy and has a higher density than the lighter one [5].

Polonite® naturally contains Al2O3 and Fe2O3 but also some heavy metals like wolfram and

vanadium [5].Before the Polonite® can be used effectively it must be heated to approximately 900°C, this tilts the reaction (Eq1) towards the right side [5]. This is

advantageous as CaO is more reactive in a water solution [6]. Heavy opoka requires slightly higher heating temperature, approximately 1000°C, as the opoka is heated both the porosity, pH and sorption capacity increases [5].

𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶3 → 𝐶𝐶𝐶𝐶𝐶𝐶 + 𝐶𝐶𝐶𝐶2 (Eq1)

After the heat treatment it is milled gently since it is desirable to have grains and not powder as it could clog the filtration systems [7]. Depending on grain size and retention time the amount of P sorption varies, but to achieve a large amount of P sorption the level of pH needs to be high as it affects the bonding mechanisms to the Polonite® [6]. Wastewaters that keep pH down can therefore interfere with the sorption capacity of P.

2.4 Reaction mechanisms

4 2.4.2 Surface Precipitation

Surface precipitation occurs when the solubility threshold of a solid phase has been exceeded. When larger amounts of material are dissolved into a solution than the solvent can dissolve under normal circumstances, the solution is called supersaturated. If it starts to stabilize towards the normal circumstances, for example at atmospheric pressure, precipitation of the solid phase begins. The surface precipitation is a three-dimensional arrangement of the molecules which composition and structure depends on the host [6].

2.4.3 Ion-exchange

Ion-exchange is the interaction between ionic species in a solid and aqueous solution. The ions in the water are removed by an electrostatic attraction, every ion removed is replaced with a similar charged ion from the solid. [6]

2.5 Surface finishing

Due to high temperatures during hot rolling and annealing alloy elements from the steel matrix diffuses to the surface where they react with oxygen in the air, this leads to the

forming of oxide layers. They are most commonly divided into two different layers, the outer and inner layer. The outer layer consists of basic oxides while the inner layer consists of more complex ones. The diffusion of the alloy elements to the surface causes a chromium depleted zone beneath the oxide layers. The steels properties decreases to deficient levels due to the chromium depleted zone. Avoiding this zone and the oxide layers to get worked into the metal in further process steps is essential. The layers are therefore removed by a process called pickling. The most common pickling methods are acid and electrolytic pickling [16].

2.6 Pickling process

Electrolytic pickling is a process usually performed after the steel is annealed. The steel is then descaled mechanically, followed by electrolytic pickling where it is submerged in an electrolyte and a current is applied [17]. These steps are performed to ease the acid pickling. The oxide layer is modified electrochemically by the electrolyte in order to increase the solubility in the acid pickling solution. This is performed in sodium sulfate (Na2SO4) which is

slightly acidic. The metal oxides are oxidized to metal ions during exposure of the electrolyte. The ions are then dissolved into the electrolyte [17].

While the metallic phase is dissolved, the oxide layers surface are undercut and removed [18]. During the treatment, trivalent chromium (Cr3+) is oxidized to hexavalent chromium (Cr6+) which is highly toxic. Before starting the acid pickling process, hexavalent chromium is reduced to the less toxic trivalent chromium, by using sodium sulfate (Na2SO4) as reducing

5 2.6.2 Acid pickling

Acid pickling is the last step in the pickling process. The steel is submerged in acid and thereby the oxide layers are removed. Sulfuric acid (H2SO4) or an acid mix consisting of

nitric acid (HNO3) and fluoric acid (HF) are the most commonly used. Depending on the acid

used, the process is either called acid or mixed acid pickling. In the stainless steelmaking production, mixed acids are solely used [17].

2.6.3 Mixed acid pickling

The mixed acid consisting of nitric acid and fluoric acid depletes the oxides on the steels surface. The depleted oxides then merge into particles that create sludge in the bath. The acid also dissolves Fe, Ni and Cr from the steel into the bath. The two acids have different tasks in the pickling bath. The main task of HNO3 is to remove the chromium depleted zone and

oxide layer. Because of its strong oxidation properties it will also precipitate metal ions into the bath. As the metal concentration rises the nitric acid gets consumed. New oxide layers do not form because of the HF being present. Furthermore, fluorides form compounds with the metal ions in the acid, which will precipitate as sludge in the bath. HF also provides hydrogen ions (H+) which contribute to the formation of new HNO3 [17].

2.7 Neutralization of wastewater

The neutralization process is the most commonly utilized method for treatment of wastewater from the pickling process [17]. It is used to increase the pH of the wastewater which then leads to precipitation of metal hydroxides due to their low solubility in water. The solubility of the metal hydroxides is strongly pH dependent. Every metal has its own optimal pH interval where it most efficiently precipitates, but this interval is often very narrow. Slaked lime (Ca(OH)2) is often used to increase the pH of the wastewaters. Sludge is then formed

containing a high metal concentration, where fluorine ion in the solution can be used as slag formers. Fluorine ions form CaF2 together with the calcium ions from the lime, which also

precipitates from the water [17].

2.8 Inductively Coupled Plasma

The Inductively Coupled Plasma was introduced in 1964 and is used for the excitation of elements within a material [19]. After the excitation, various analysis methods can be used, such as Atomic Emission Spectrometry (AES) or mass spectrometry (MS). The plasma advantages over similar heating methods are both the wide range of composition it can excite, and that it can affect multiple elements simultaneously. Argon (Ar) is most commonly used to produce the plasma. The heat is usually between 6000 and 10000 K, which is enough to ionize most elements and to break the molecular bindings. Both Helium (He) and Nitrogen (N) can be used as well. ICP operates by leading the gas to an induction coil, which is induced by a radio frequency generator to transmit the energy to the gas [20].

6

Ar. The excited electrons will collide with other Ar molecules, creating a steady stream of ions, resulting in a raise of temperature, while plasma is created. [20] The material is added as a solution in small droplets that combined with the Ar form an aerosol, which is sprayed into the plasma from a torch [21]. The samples are then examined with AES or MS.

2.9 Conductivity

This parameter measures how well electrical current can travel through a medium, in this case water [22]. Conductivity gives a general indication of the quality of the water, where low conductivity is desirable. If the conductivity changes unexpectedly during a process it could signal that some properties inside the water are changing as well. The method is simple to use and gives results instantly. The variation of ion content can cause the variation of conductivity, were ions such as H+, OH- and phosphate contribute to higher conductivity [23].

2.10 Environmental aspects

The P refinement is fundamental to close the loop of slag and sediments after neutralization, as shown in fig. 1. Polonite® would be used as a precursor to the neutralization procedure, allowing the sediments to be brought back to the steelmaking process instead of going into landfills. The amount of material waste going to landfills needs to be reduced by minimizing the byproducts and by retrieving what can be recirculated. Polonite® containing high

amounts of P have been tested for agriculture use, but since the wastewaters tested contains various different elements, its usage is not confirmed. If Polonite® cannot be recycled for external use it does not solve the landfilling issue.

7

8

3 Experiment

A batch and a column test were performed, the batch test to estimate Polonite’s® adsorption abilities by finding its equilibrium with the wastewater. The column test were the main test regarding its ability to remove Phosphorus, P, in a constant water flow. The flow of

wastewater was used to determine Polonite’s® kinetic properties, as for most neutralization plants the process occur in a kinetic state.

In the batch test Erlenmeyer flasks containing a set amount of wastewater and varying amounts of Polonite® were used find the equilibrium. The column test consisted of three columns filled with the same amount of Polonite®, through these columns wastewater was pumped through with a constant flow. From this its kinetic properties could be evaluated.

3.1 Equipment

• Shaking table, from ”Stålprodukter” • Acid washed bottles

• Polonite®

• Parafilm trademark, M

• Erlenmeyer flasks/Conical laboratory flask • Pipette

• Sieves

• Three digit scale

• 3 sewage pipes in polypropylene, 100 [cm] diameter, approximately 70 [cm] long • Timer

• Ismatec ECOLINE VC-MS/CA8-6 (pumping machine) • Untreated wastewater from Sandvik

• PHM 95 pH/10N Meter, Radiometercopenhagen. • Orion 115A+ (Thermo)

• Cotton

• Plastic bottles

• Tubes, 2 [mm] inner diameter.

3.2 Batch test

As Polonite® is processed bedrock, the composition will vary. To test the maximum sorption of P four series of Erlenmeyer flasks (Conical laboratory flasks) were prepared. The

Polonite® was delivered in buckets with varying grain sizes, in this experiment the size was set to 2-4 [mm] and sieves were used to filter out the requested particle size.

9

cylinder. The top was sealed with Parafilm, this to prevent evaporation, then put on a shaking table for 24 hours to find the chemical equilibrium between Polonite® and the wastewater, see fig. 2. If there was no decrease of the pH probably no P would have been adsorbed by the Polonite®. Therefore a pH test was conducted.

Table 1. Display of how the Erlenmeyer flasks were prepared with various amounts of

Polonite®, three of each amount. The first flask of each is referred to as the series letter and a “1”, for example the first flask in Series D is called D1, second is referred to as “2” and the third as “3”. Series 1 2 3 A[g] 5 5.01 5 B[g] 10 10 9.99 C[g] 15 15 15 D[g] 0 0 0

10

3.3 Column test

The three columns used were sealed in the bottom with a rubber plug both to ensure it was waterproof and to retain the Polonite®. Since leakage during the experiment would ruin the flow of water in the columns, altering the results, Polonite® particles were sieved in steps to achieve the grain size of 2-4 [mm]. First, a filter with the grid size of 5.8 [mm] was used, then 4.0 [mm] and last 2.0 [mm]. At the bottom a container was installed to catch any smaller particles. The grains caught between the 4 [mm] and 2 [mm] layer are the ones used.

In this test Polonite’s® kinetic properties was tested, to see how it reacts and compares to the previous equilibrium test. For this experiment, 590 g of Polonite® in each column was weighed and carefully poured into the columns, for exact values see table 2. This amount was estimated from the results from the batch test, calculation in appendix D. The plugs were equipped with an outlet hole for the treated water to pass through. To make sure it would not clog a filter was installed over it. The outlet was connected to a plastic tube, which was tilted back up toward the top of the column to prevent the water to flow out too fast and thereby ruin the saturated condition of wastewater. The natural hydrostatic pressure was set up to be the operating force of the outflow of wastewater. To prevent the siphon effect air release valves were set up on the top of the outlet tubes. After the air valves, treated water ran down the tubes toward a designated container, this to be able to measure and compare the columns. To ease the transportation of the purified water, the containers were put below the table upon which the stance stood.

Table 2. The amount of Polonite® and initial wastewater poured into the columns.

Column 1 2 3

Polonite® [g] 591.89 581.32 591.64

Initial wastewater [ml] 500 520 530

Fig. 3 shows how the three columns were set up during the experiment, as vertically as possible, this to prevent preferential flow in the columns. To pump the water to the top of each column, an Ismatec ECOLINE pumping machine was used to transport the wastewater from the original plastic container to the three columns. As the wastewater contained

11

Fig. 3. Setup for the column test. On the left the setup is shown from the side. The three columns, pumping machine and the wastewater container displayed. To the right is an

overview of the columns displaying a saturated condition, the white part of the tubes is the air release valves.

12

Table 3. List of what time of the day each of the periods relates to, with the responding amount of wastewater that had flowed through each column.

13

4 Results

4.1 Batch test

From the initial values there is an increase of pH in all samples, see fig.4. The flasks with least amount of Polonite® (5 [g]), the A-series, resulted in the lowest pH. With increasing amount of Polonite® the pH increased, where the C-series (15 [g]) had the highest with an average of 12.48, B-series had 11.99 and A-series 7.78. The initial average was about 1.56. The largest difference in pH, within the series, was in the A-series with 0.81.

Fig. 4. Display of the pH after the batch test. On the y-axis the specific sample and on the x-axis the corresponding pH.

The samples derived from the D-series set the initial value of P in untreated wastewaters with an average value of approximately 2.67 [mg/l] as shown in fig. 5. In C-series the results from C1 vary with a factor of 10 from the other two, therefore it was considered an outlier. The average amount of P in series C is 0.05 [mg/l]. With C1 excluded, this is approximately 1.9% of the average amount of the initial value (D-series).

From the received results for A and B-series, the amount of P had been reduced even more effectively than the C-series. With over 98% reduction in both cases, the series with lower amounts of Polonite® (A and B) have the lowest values of P. The average in A-series were 0.032 [mg/L] and 0.03 in B-series. In table 1 the different series are shown.

14

Fig. 5. Plot of the amount of P in the wastewater from the batch test. On the x-axis the amount of P and the on y-axis the specific sample. The red bars are the variation of the tests.

The untreated wastewater contained the most amount of F, with an average of 1300 [mg/L], see fig. 6. The largest reduction of F was in the sample with least amount of Polonite®, and the amount of F increased with increasing amount of Polonite®. The amounts did not differ too much between A-C-Series, the average amount of F in A-series was 5.3 [mg/L] and 6.07 [mg/L] in the B-series. In the C-series the average was 7.13 [mg/L], this is about 0.55% of the initial value (D-series), all the series had over 99% removal of F.

15

4.2 Column test

Compared to the initial value there was a drastic drop in conductivity for the treated water, the first 6200 [ml] of used wastewater had an overall declining conductivity trend in all columns, (P1-P6) see fig. 7. It is noticeable how after the 6200 [ml] the conductivity drops consistently for all three columns after a phase of pseudo stability. Column 1 (Col.1) had a lower drop than Column 2 and 3 (Col.2 and Col.3) from approximately 6200-7400 [ml] of total water used. These drops were quite significant as the conductivity of Col.2 dropped by 49.5% and Col.3 by 45.4%. Col.1 dropped by 25.3% during the same period. The

conductivity in the last 2 samples seemed to be similar.

Fig. 7. Plot of conductivity for the untreated wastewater and of the water collected from the columns after the designated periods, see table 3. Some of the highest and lowest values are displayed for reference. The conductivity is displayed on the y-axis and the total amount of used wastewater on the x-axis.

The pH of the first water collected from all columns greatly differed from the initial value of approximately 1.6, thereafter the pH did not vary much before over 6000 [ml] had been pumped through, see fig. 8. Between approximately 6200-7300 [ml] of total amount of water pumped through the columns, Col.1 had a decrease of pH from 12.59 to 12.17. Col.2 dropped from 12.57 to 8.4 and Col.3 from 12.49 to 4.3 after the same amount of water load.

The overall declining trend of Col.1 and Col.2 was greater than Col 3, with the largest decline in Col.1. In Col.3 there was an increase of pH when approximately 4900-6200 [ml] had been used, this behavior of a significant increase which was exclusive to that column.

16

17

5 Discussion

For the batch test there was an overall decrease of P in all samples that had been treated with Polonite®, see fig. 5. This indicated that P could be successfully removed from this type of wastewater using Polonite®. Regarding the amounts of Polonite® used and the reduction of P there seemed to be an unusual correlation as the Series (A and B) with least amount of Polonite® showed the highest P removal capacity. This phenomenon is in contrast with earlier studies that have successfully shown that there should be a direct correlation between amount of Polonite® and the amount of removed P [1]. This was the case even if the value of C1 (15 [g]), which was considered an outlier, was excluded from the average of the C-series. The amounts of F, in fig. 6 had a similar trend, as it had the lowest amount in series A (5 [g]), increased with increasing amount of Polonite®. The high acidity of the wastewater used for this experiment was unprecedented, which could lead to yet unforeseen results.

The samples were shaken on the same shaking table for the same amount of time. 24 hours was considered to be enough time to reach equilibrium between the water and Polonite®. Therefore, kinetics is not considered as a factor that might affect the results. There was also no confirmation of the wastewater containing homogeneous distribution of elements, and despite the fact that the pH did not vary much, there could be other variations in composition affecting the results since no full screen ICP analysis was made. The difference in amounts of P and F in the series could also be due to particles being filtered out before the ICP was conducted at the lab, yet no detailed information of their procedures have been received. The low amount of F left in the water made the use of Polonite® questionable for this specific application, as it did not satisfy Sandvik’s demands.

For the most part of the column test, when approximately 6200 [ml] of wastewater had run through the columns, the pH value barely differed from 12.6 see fig. 8. This could be due to too much Polonite® in the columns. If less was used the pH would have probably decreased steadily as the Polonite® was getting exhausted over time. The amount of Polonite® used in the columns was decided based on the results from the batch test, although this choice did not give the expected results. A reason could be that the Polonite® interacted with the wastewater differently in the two tests, as one was equilibrium based and the other kinetic. The batch test was conducted to find an approximate equilibrium, while the column test was used to

understand how Polonite® reacts in a constant flow of water.

As the water flowed down the columns, the Polonite® should have gotten more and more exhausted leading to a less P adsorption from the wastewater. An odd behavior can be noticed in both Col.2 and 3 where after 6000 [ml] of wastewater there is a sudden drop in the pH levels. This phenomenon could be explained by the formation of a preferential flow within the columns. The water traveled downwards through a specific path created by macropores, limiting the surface area available to react with the water. The pace of the water in the

18

the higher constant hydraulic load. In Col.1 there is only a slight decrease of pH after the same water load as in Col.2 and 3, which corresponds better with previous results regarding Polonite’s® exhaustion rate. As no ICP test regarding the column test had been received, no correlation could be established regarding the pH value and removal abilities of P and F.

During the column test there was an overall decrease of conductivity see fig. 7, this could indicate less metallic ions present in the water during the later part of the experiment According to an earlier study there is a correlation between how H+ and OH- contribute to conductivity depending on the pH of the water [23]. The lower the pH the lower the

contribution, this applies to phosphate contribution as well. A pH of 7.2 is a general threshold for H2PO4- converting to HPO42- and this also resulting in a higher conductivity with higher

pH [23]. The conductivity in both Col.2 and Col.3 had a similar result in approximately the last 1000 [ml], the pH differed greatly as Col.2 had 10.7 while Col.3 had 4.306.

The initial water had a conductivity of 16 [mS/cm] and a pH value of about 1.6. The results indicated a decrease of the conductivity, while the pH value increased heavily in the early stage of the column test. Conductivity and pH has a direct proportionality, although the results showed the opposite in the beginning. Particles in the untreated water could be filtered out by the Polonite®, Polonite® in this case acts as a mechanical filter. As the wastewater contains a lot of particles, this could affect the conductivity greatly. Ion exchange also affects conductivity, probably not as drastically. Polonite® exchanges ions with similar charge, leading to a smaller net change in conductivity. If the particles were the major contributor to the high initial conductivity, removing them would explain the decrease. These values were then quite stable until about 6200 [ml] of total water had been used.

If the macropores forced the water to pass through a preferential path, limiting the reaction with Polonite®, hence the particles still could have been filtered out. The reason that both the pH and conductivity decreased could be that the particles, in this case regarded as the most influential contributor to the conductivity, were removed and therefore the pH and

conductivity’s direct proportionality occurred in fig. 7 and 8. The pH would then drop toward the initial value of about 1.6 while the low value affects the conductivity, decreasing it as well. Depending on the size of the macropores, the amount of water streaming through the preferential flow would differ, thus explaining the different declines of pH in the columns. The largest preferential flow would lead to the lowest pH value, due to the higher exhaustion rate of the Polonite®.

The ICP analysis of the column test has not yet been received, a discussion of these results can therefore not be made.

5.1 Environmental aspects

19

the world. If the landfill problem is not fixed, the Phosphorus supplies in the world will diminish, less food will be produced due to the lack of fertilizers. This will lead to both an increase of starvation and higher food prices, which also limits who can afford to buy it.

Many elements that are put into landfills does not belong there, some are recyclable but may be contaminated. The energy needed to manufacture or produce elements is usually higher than the energy that is needed to recycle it, putting material in landfill is therefore a great energy loss as well as a material loss, due to the limited supplies. This makes excessive landfilling highly unethical and should be avoided.

5.2 Limitations

The economic benefits or costs of the usage of Polonite® are not included. This because there are numerous costs impossible to estimate in this study. For example the cost of landfilling sediments and slag is hard to evaluate therefore any approximations would defeat the purpose of an economic evaluation. There was no chemical analysis regarding the composition of the exhausted Polonite®, therefore no confirmation of its use as a fertilizer can be obtained. Instead a full screen analysis of the composition of the water was sent but no results received. All the water that had passed through the columns, except for the amounts used for sampling, was mixed to create an average of the treated water. Its composition was then compared to the initial water, with the assumption of Polonite® containing the elements purified from the water during the experiment. Only one type of grain size of Polonite® was used, all

20

6 Conclusions

Polonite® can be used to reduce the amount of Phosphorus to the target level of 1-2 [mg/L]. Polonite® is capable of binding with other elements as can be seen by the incidental Fluorine removal from the wastewater. The amount of Polonite® used in the column test was

overestimated. Using the batch test to estimate the amount of Polonite® for the column test was a bit too rough of an approximation.

21

7 Recommendations

22

8 Acknowledgements

We would like to thank the people who have made this project possible:

Olle Sundqvist, R&D Expert in Metallurgy at Sandvik AB, our supervisor at Sandvik, who invited us to Sandvik in Sandviken for a field trip, making the problem more understandable.

Mattia de Colle, PhD Student on Dep. Of Materials Science and Engineering KTH, main supervisor, for helping, encouraging and giving us relevant feedback making this project successful.

Professor Pär Jönsson, Division of Applied Process Metallurgy KTH, co-supervisor, for making this project possible as well as all the encouraging words along the way.

Agnieszka Renman, Division of Water and Environmental Engineering, Laboratory supervisor, for letting us use the laboratory and guiding us through the equipment’s being used.

23

9 References

1. World Steel Association. Water Management in the Steel Industry [Internet]. Bryssel: World Steel Association; 2011. [cited date 2018 Apr 2] Available from: https://www.worldsteel.org/en/dam/jcr:f7594c5f-9250-4eb3-aa10-48cba3e3b213/Water+Management+Position+Paper+2015.pdf

2. Springco Metal Coating. Metal Pickling [Internet]. Cleveland: Sprinco Metal Coating; c2015 [cited 2018 Apr 2]. Available from: http://www.springco-coatings.com/services/metal-pickling.html

3. Sandvik AB. Materialdatablad [Internet]. Sandviken: Sandvik Materials Technology; c2018 [cited 2018 Apr 2]. Available from: https://www.materials.sandvik/se/materialcenter/materialdatablad/

4. Ispat Guru. Phosphorus in Steels [Internet]. India: Ispat Guru; 2014 [cited 2018 Apr 2]. Available from:

http://ispatguru.com/phosphorus-in-steels/

5. Brogowski Z, Renman G. Characterization of Opoka as a Basis for its Use in Wastewater Treatment. Polish Journal of Environmental Studies 2004;19(1):15-20

6. Nelin C. Evaluation of using fine grain size Polonite® as sorbent for retaining phosphorus from wastewater. [master’s thesis on the Internet]. Lund Institute of Technology; 2008 [May]. Available from:

http://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=1320871&fileOId=1320872

7. Renman A, Renman G. Long-term phosphate removal by the calcium-silicate material Polonite® in wastewater filtration systems. Chemosphere. 2010 Mar;79:659-64

8. Encyclopaedia Britannica [Internet]. Encyclopaedia Britannica Inc; 1998. Adsorption. [cited 2018 Apr 6]. Available from: https://www.britannica.com/science/adsorption

9. Emsley J. Nature’s Building Blocks. New ed. Oxford University Press; 2011

10. Otai Special Steel. 21 Chemical Elements and Effects on Steel Mechanical Properties [Internet]. Dongguan, Guangdong Province: Otai Special Steel; 2015 [cited date 2018 Apr 6]. Available from:

https://www.astmsteel.com/steel-knowledge/chemical-elements-and-effects-mechanical-properties/

11. Agriculture Solutions Inc. Phosphorus (P) [Internet]. Sebringville, Ontario: Agriculture Solutions Inc; c2012 [cited date 2018 Apr 6]. Available from: https://www.agriculturesolutions.ca/phosphorus-overview

12. Pedersen K, Elding LI, Björn LO, Erlandsson U. fosfor. In: Nationalencyklopedin [Internet]. Malmö: Nationalencyklopedin; [unknown date]. [cited date 2018 Apr 6]. Available from:

https://www.ne.se/uppslagsverk/encyklopedi/l%C3%A5ng/fosfor

13. Karbowniczek M, Kawecka-Cebula E, Reichel J. Investigations of the Dephosphorization of Liquid Iron Solution Containing Chromium and Nickel. Metallurg and Mat Tran B. 2012 June;43B:554-61

14. Earth Sciences [Internet]. Minneapolis-Saint Paul: University of Minnesota; c2012-. Fluorite [cited date 2018 Apr 8]. Available from: https://www.esci.umn.edu/courses/1001/minerals/fluorite.shtml

15. Wilkes G. Hydrofluoric Acid Burns [Internet]. New York: Medscape; 2016. [cited date 2018 Apr 8]. Available from: https://emedicine.medscape.com/article/773304-overview

24

Technology; 2017 [cited date 2018 Apr 4]. Available from: http://www.diva-portal.org/smash/get/diva2:1113907/FULLTEXT01.pdf

17. Dahlgren L. Treatment of Spent Pickling Acid from Stainless Steel Production [master thesis on the Internet]. Stockholm: KTH Royal Institute of Technology; 2010 [cited date 2018 Apr 5]. Available from:

http://www.diva-portal.org/smash/get/diva2:473369/fulltext01

18. Li LF, Caenen P, Jiang MF. Electrolytic pickling of the oxide layer on hot rolled 304 stainless steel in sodium sulphate. Corrosion Science. 2008 July;50:2824-30

19. Ulrich N. Atomic Emission and Absorption. Encyclopedia of Food Sciences and Nutrition. 2003;2:5441-7 20. HiQ Linde Gas. Inductively coupled plasma [Internet]. München: HiQ Linde Gas; c2018 [cited date 2018 Apr 12]. Available from: http://hiq.linde-gas.com/en/analytical_methods/inductively_coupled_plasma.html

21. Laboratory Testing Inc. ICP Analysis [Internet]. Philadelphia: Laboratory Testing Inc; c2018 [cited date 2018 Apr 12]. Available from: https://www.labtesting.com/services/materials-testing/chemical-analysis/icp-analysis/

22. Lenntech BV. Water Conductivity. Delfgauw:Lenntech BV; c2018 [cited date 2018 Apr 12]. Available from: https://www.lenntech.com/applications/ultrapure/conductivity/water-conductivity.htm

Appendix

Appendix A. Table of the measured conductivity for the column test.

Conductivity C1 C2 C3 P1 8.69 8.87 8.65 P2 9.89 9.61 9.61 P3 9.31 8.86 9.06 P4 9.72 9.34 9.14 P5 9.52 9.46 8.98 P6 8.88 8.76 7.62 P6 N 9.24 8.74 8.18 P7 6.9 4.5 4.53 P8 6.41 4.45 4.52

Appendix B. Table of the measured pH in the column test.

Period Column 1 Column 2 Column 3 Initial value

Appendix C. Table of the measured amounts of P and F for the batch test, variation included.

Sample P [mg/L] variation F variation2

A1 0.025 0.005 5.4 0.81 A2 0.038 0.005 5.4 0.81 A3 0.034 0.005 5.1 0.76 B1 0.05 0.005 7.3 1.1 B2 0.025 0.005 5.5 0.83 B3 0.016 0.005 5.4 0.81 C1 0.5 0.05 7.5 1.1 C2 0.05 0.005 6.5 0.98 C3 0.05 0.005 7.4 1.1 D1 2.6 0.26 1400 210 D2 2.7 0.27 1300 200 D3 2.7 0.27 1200 180

Appendix D. Calculation of approximated use of Polonite® for the column test.

6000 𝑚𝑚𝑚𝑚