Modern cement production with chalk

As regards conventional carbonate rock (limestone, marl), problems of quarrying and preparation are only dealt with in passing. Meanwhile there are proven standard methods for the majority of these widespread cement raw materials, and it is easy to compile a practicable quarrying and preparation concept based on these methods. Chalk, however, requires an approach taking into account the peculiarities of this material and of the specific deposit when selecting the process. Many methods and equipment for quarrying, handling and storage, working well in other cases, fail with this moist, soft and sticky carbonate rock. Special solutions are needed.

Chalk is a soft (< 5 MPa), highly dispersed (D50 < 5 µm) and moist (15 – 20  % of H₂O) limestone. It mainly consists of fossil, microscopically small calcite laminae (coccolithes), which can only be detected under an SEM (Fig. 1). The loose fabric is a consequence of an incomplete diagenesis. Flint layers may occur in certain horizons. Pyrite nodules and macrofossils (belemnites, echinites and others) also accompany the chalk.

When being dried, the quarry-wet chalk becomes a bulk solid with a high specific surface (~  10 000 cm²/g). When being mechanically stressed, the quarry-wet chalk begins to clot. The  water not contained in capillaries rises to the surface, moistens it and, consequently, generates a certain stickiness. Quarry-wet chalk is changed to slurry when being pressurized and when water is added. Usually the CaCO₃ content varies between 85 and 99  %. Within a deposit the range of variation is usually ­between 5  % and 8  %. The carbonate purity has a great ­influence on the rheological behaviour of the filter cake, which is used as kiln feed in some processes. Due to its whiteness and high specific surface, white chalk is a valuable filler for many applications (e.  g. in the paper industry).

“The transition from the wet and semi-wet process to the dry and semi-dry process with all its energetic and operational advantages has a great importance for many cement factories all over the world up to now”. This statement of Palle ­Grydgaard [1] is undoubtedly correct and applies to the majority of the cement factories all over the world. But there is no rule ­without an exception. “If the raw material is very fine and moist, such as chalk, or if the chloride content exceeds an allowable limit, the wet preparation still has a right to exist today”. ­Already in the 1980s FLSmidth introduced the so-called “compact kiln system” for such moist raw materials (Fig. 2).

This process quickly proved a success and other plant manufacturers (Polysius, KHD) soon followed with various modifications. The waste heat of the kiln system is used as a heat source for raw material drying. In order to have sufficient drying heat after the cyclone preheater (8), the effectiveness of the latter has to be restricted by reducing the number of cyclone stages. After leaving the top cyclone stage (8), the now sufficient hot kiln process gases are passed to the impact mill (5) where they carry along the moist raw material fed there (6) into the flash dryer (4). The moist raw material (6) is completely dried. The dried raw material is separated from the gas flow (7) in the cyclone separators (3) and is fed to the cyclone preheater (8). Portland cement clinker (12) is generated after passing the calciner (9), the rotary kiln (10) and the cooler (11). The process gases are passed to the filter (2). The required system draught is generated via a waste gas fan (1). The impact mill (5) with connected flash dryer (4) continuously runs in interconnected operation with the kiln. The material is not ground in the impact mill (5). There the moist feed is distributed over the entire cross-section of the passing flow of drying gas. Therefore, the raw material fed must already have a sufficient fineness.                 

Figure 3 shows the connection between the raw material moisture and the specific heat consumption of compact kiln systems with calciners. The higher the feed moisture, the more drying heat is required and the less cyclone stages may be installed. With more than 25  % of moisture (slurry), only a two-stage preheater can be used. In the range of 18 –23  % of moisture a three-stage preheater will do, and with 10  –18  % even a four-stage preheater will provide sufficient heat.

The drying system “impact mill/flash dryer” has some remarkable properties. It not only acts as dryer but is also able to homogenize raw mixes and to wash gas (scrubber). Flash dryers are used with and without installed separators (static and dynamic).

For the time being seven different processes for the manufacture of Portland clinker are known (Fig. 4). Five of them (A-E) are used in the wet, semi-wet and semi-dry processes [2]. At the end of 2010, one kiln line at Mordov Cement in Russia operating according to the dry process F is to be commissioned (Mordov 3). A project study for a semi-wet process G was prepared in Lägerdorf. The classical wet process A and the LEPOL^© process B will not be further considered. The processes C to G, however, will be presented and evaluated in detail including their advantages and disadvantages. 

The first compact kiln was built by FLS in 1982 at the Southam works (UK), which, in the meantime has already been shut down. Aalborg (DK) and a short time later Rugby (UK) followed in the early 1990s. Since then both plants have been run successfully. The locations Aalborg and Rugby have a long history. Both use the wet preparation of chalk untill today. What were the reasons not to change over to a dry process, differing from the general trend of the industry?

Almost the entire chalk deposit of Aalborg is located in the natural ground water. Furthermore, the existing wet kilns should be kept for the production of white cement. Rugby gets it chalk from the Kensworth quarry located at a distance of 95 km. Hydraulic conveying proved to be the most economic ­alternative for the transport of the chalk from the Kensworth quarry to the factory. The pipeline has been in operation since the middle of the 1960s without any major problems and was updated just a short time ago [3].

The kiln waste gases of a two-stage preheater are used to dry the cement raw slurry with a water content of 34–36  % in the flash dryer. A filtration to mechanically dewater the cement raw slurry previously was not taken into account, firstly to avoid additional capital spending and secondly because the chalk relatively poor in chloride does not require any washing out of chloride.

Lägerdorf is producing according to this process as well as at Mikhailovka (RUS, Sebriakovsk Cement) and Mordov Cement (RUS, project Mordov 2).

The introduction of the LEPOL^© technology and finally also the high chloride input into the chalk (up to >  0.3  g Cl^-/kg chalk related to dry matter) due to penetrating pit water forced Lägerdorf to use the filtration technique already in the 1960s. Due to this hydraulic peculiarity at the Lägerdorf chalk quarry, the already existing filtration system (Fig. 5) was kept when installing the new kiln line 11 in 1995. When washing with water poor in chloride (<  200  mg Cl^-/l) and subsequent filtration, rates of washing out salt of more than 60  % [4] from the chalk have been achieved. The relatively low specific heat consumption of D is actually only a welcome side effect. From the point of view of process engineering, the quarry situation was decisive as regards the selection of the preparation concept.

Difficulties and missing experience as regards the operation of chamber filtration, recently led to process modifications at Sebriakovsk Cement. Thus, today part of the raw material is injected into the flash dryer as slurry (about 15  % related to dry matter). The remaining part is still fed as filter cake. Thus, one could say that there a kind of hybrid of the processes C and D is used.

Dosing problems regarding the filter cake led to considerable delays in the project Mordov 2 in 2008. There the washing-out effect of filtration obviously does not play a decisive role.

Many years of experience show that the higher the degree of purity (% of CaCO₃) of the chalk the more difficult it is to handle the filter cake produced in the chamber filter press. This, of course, has negative effects on the storage and transport behaviour. As opposed to the chalk of the French Dannes (85  % of CaCO₃), where they installed a LEPOL^© kiln, the chalk from Lägerdorf has a considerably higher content of carbonate (up to 98  %). This represents the most exacting individual case as regards (chamber) filtration and filter cake.

In now more than 40 years of chamber filtration for chalk-based cement raw slurries sufficient experience has been gained at Lägerdorf as regards the production and handling of filter cake to ensure a stable kiln operation. In addition, the existence of a filtration system offered the possibility of an unusual way to make use of Cl bypass dusts. For the time being, about 50 % of the daily yielded bypass dust are used as substitute for milk of lime for the conditioning of the filter cake. Currently there are activities to set up a special filtration system to wash the remaining bypass dust and return it to the process.

This process is very popular. For the first time it was tested by FLS in Chelm (PL). It differs from all the other processes by the absence of the chalk slurry production.

Shovels dislodge large lumps of chalk from the face in the quarry of Chelm. Trains transport the lumpy chalk from the quarry to the raw material preparation section. After unloading of the wagons, the chalk passes two roofed crusher stations where it is comminuted to pieces of max. 50 mm edge length. This crushed chalk is passed to a large, also roofed longitudinal store (Fig. 6). The stored chalk is reclaimed by means of Bedeschi reclaimers (Fig. 7) and is fed to a belt conveyor. Via proportioning boxes the chalk is passed directly into the flash dryer together with additional marl components according to the raw material mixture. At first sight this process seems to be unrivalled lean compared to all the other processes. However, this is only partially true on closer examination:

– To be stored in stockpile, the chalk has to be extracted in coarse lumps. Fine chalk would very soon become sticky in the open air, which would make the storage in longitudinal stores difficult. This hinders the use of effective extraction units in the chalk quarries (e.  g. bucket wheel excavators + continuous conveyors) because these extract the material in small sizes. If the quarry is additionally near the ground water, railway or truck transport on the chalk will also be a challenge. A large number of employees is required to maintain such transport systems.

– An operation in Chelm is virtually impossible in case of a long lasting frost. Frozen chalk lumps cannot be dried completely in the flash dryer since the waste heat of the kiln after the third cyclone stage is not sufficient for thawing and drying of the chalk.

– The natural moisture of the chalk varies between 21 and 28  % depending on the weather conditions. Process E only allows limited control of the feed moisture, which has an unfavourable effect on the drying and burning processes.

– The “dry” storage of chalk does not allow a safe separation of foreign bodies (flints, macrofossils, excavator teeth etc.). The rotor of the impact mill in Chelm does not survive a kiln campaign. It only runs for some months. The kiln must be stopped for a rotor change.

Process E, convincing at first sight, requires an exact examination of the overall process, starting at the chalk quarry. Some seeming advantages could quickly be changed into serious disadvantages if not all links in the production chain are included in the analysis. It should also be mentioned here that process E does not belong to the processes with the lowest specific heat consumption although it is a semi-dry process.

If there is enough useable waste heat from other processes, it can be used for a dry preparation of chalk. For the project
Mordov 3 it is planned to dry and grind the chalk and the clay components in a vertical mill. Then the dried raw meal is to be fed to a conventional preheater kiln (not a compact kiln system) with five cyclone stages (F). In this case no impact mill including flash dryer will be required. The specific heat consumption is expected to be somewhat in the order of (800 + 3000) kJ/kg clinker. Process F, however, only makes sense if there is a safe access to external waste heat (800 kJ/kg clinker). Thus, the kiln system is directly connected with this secondary process (source of waste heat).

A further development of the currently used process D for kiln 11 has been investigated at Lägerdorf. Instead of mechanical dewatering of slurries of chalk/clay and chalk correcting materials (42–45  % of H₂O) down to a residual moisture content of 19–21 % with chamber filter presses, pure chalk slurry (40  % of H₂O) is to be dewatered to a residual moisture content of 13 –14  % with membrane filter presses in the new process. Then sufficiently fine correcting material and the chalk filter cake will be fed separately to a compact kiln system. The low residual moisture of the filter cake will allow the addition of a fourth cyclone stage.

Thus, a specific heat consumption of about 3200 kJ/kg clinker will be within reach. This process G would have the lowest heat consumption value of all chalk processes examined so far.   

A peculiarity of the process is that the final raw mix will only be formed in the flash dryer by combining individual raw material flows (chalk, fly ash, finely ground ores and sand). Already today (D) two material flows come together in the flash dryer at Lägerdorf (filter cake = chalk + ore + sand and the fly ash). In the process G more than three raw material flows will be fed to the flash dryer. Thus, the material flexibility of the burning process will be expanded. Without previous changes of the slurry mixture, the clinker chemistry can be changed later virtually during burning. The change from one clinker quality to the other (e.g. between OPC clinker and sulphate-resisting clinker) can be carried out within a short time.

After this survey of all cement-based processes of clinker production, now the peculiarities of the quarrying and preparation of chalk will be examined in more detail.

There is a close connection between quarrying, transport and storage of chalk, on the one hand, and the following burning process, on the other hand. Basically, soft chalk can be ­extracted with any kind of excavator [5], even scrapers and graders are used with dry conditions. However, mainly shovels, bucket wheel excavators and bucket excavators are used. This modern quarrying equipment differs as regards the excavated material: shovels break out the material from the face and the chalk is rather lumpy; bucket wheel and bucket excavators cut out small sizes.

Shovels will cause a rather discontinuous removal of the excavated material. If they still use a combination of shovel and belt conveyor (e.  g. the Omya chalk works at Rügen), a proportioning box feeder with integrated crusher will be required. By contrast, the continuously extracting bucket wheel and bucket excavators are running in combination with continuous conveyors [6]. The use of bucket wheel excavators in combination with trucks [7], however, has not gained acceptance.

As already mentioned, the fragmentation of the quarried chalk has a decisive influence on the following storage. Process E is based on a “dry” storage of lumpy chalk (Fig. 8, top line, red). If this process is chosen, the kind of the quarrying equipment is predetermined, i.e. is must be an excavator extracting lumps. Only in this way can it be ensured that the surface of the chalk, which can be wetted by precipitations, will be kept low per ton of material being conveyed. The following processing (crushing and storage) must be roofed.

The finer the chalk is broken out from the face, the specifically larger will be the wettable surface of the material being conveyed leading to an increased inclination to clotting and sticking. The use of bucket wheel and bucket excavators would complicate a later “dry” storage of chalk. Corresponding tests with material from bucket excavators were carried out at Läger-dorf in the early 1990s [8]. Process E without slurry requires lumpy chalk that has to be removed by rail or trucks.

If the material is quarried with draining, transport problems will occur. If ground water leaks out from the quarry floor or faces, transportation by railway or trucks will be a challenge as regards maintaining the conveying line. In such cases a combination of continuously quarrying bucket wheel or bucket excavators with belt conveyor systems (Fig. 8, two bottom lines) has proved a success. However, “dry” storage (as with E) will not be possible. In addition, there is no possibility in process E to remove foreign bodies from the chalk. This has negative effects on the wear behaviour of the rotor of the impact mill.

The complicated deposit conditions at Lägerdorf have already been mentioned: The entire deposit is located in the ground water. If deep water rises, salt will be put in. The horizons are interstratified with flint layers. These quarry-specific conditions alone only allowed the processes D or G for the kiln concept in Lägerdorf (Fig. 8 central line).

Due to the special hydrogeological conditions at the Limfjord (no drainig possible), the major portion of the chalk in Aalborg has to be quarried underwater (Fig. 8, bottom line). At that time only bucket excavators were available. Quarrying of small sizes and high moisture contents of the material excavated exclude “dry” storage. Only the processes C, D and G were available for selection.

The economic advantages of hydraulic handling of chalk over large distances were decisive in Rugby for the choice of the wet processing though the chalk quarry located above the ground water the “dry” storage would also have been admitted. Transportation in trucks is possible due to an effective drainage of the routes (Fig. 8, top line, black).

Experience has shown that the belt conveyors are the weakest link of the chain with the high-capacity quarrying complexes using bucket wheel or bucket excavators. They are not sufficiently flexible when quarrying chalk in quarry ranges with difficult access. The operating costs for belt conveyors are relatively high. In case of frost additional problems will occur.

Since the middle of 2009, the chalk broken out from the face of the Lägerdorf quarry is suspended in a pilot plant directly at the working place and is then pumped hydraulically as chalk slurry. This mobile slurry-forming plant for 150 t/h chalk was designed and manufactured within the framework of a joint project between Omya/VDK and Holcim (Deutschland) AG. Via a pipeline the material is removed from the quarry at a depth of 90 m and is directly conveyed to the chalk works at a distance of one and a half kilometre (Fig. 9). This new technology has some important advantages:

– high flexibility of excavator operation; at Lägerdorf an additional 30 – 40 million tons of chalk are opened up in ranges of the quarry that could not otherwise be reached

– the operating costs are considerably reduced

– theory: better preparation behaviour in case of frost.

The theory of better frost compatibility goes back to the following consideration. A bucket wheel excavator gets hold of frozen chalk only at the block edges. The chalk is largely frost-free in the face volume even with low temperatures. The excavated material is crumbled in the crusher of the mobile washing plant by adding water. Thus, the wettable surface is increased, via which the heat transfer from the washing water into the frozen parts can be intensified. The danger of freezing during the further transport in a pipeline can be more easily controlled compared to other kinds of transportation (insulation, trace heating, buried laying). Decades of experience gained with conventional wet processing allow many Russian plants to maintain the operation of the slurry tanks even in heavy frost after the material has arrived at the plant. The gap in quarrying could be closed by the new technology. It is amazing, but just the wet preparation of chalk represents the better alternative for winter operation. There are similar approaches for the quarrying of oil sand in Canada under similar climatic conditions [9].

An extraction method for chalk deposits, which are completely in the ground water, is conceivable but so far has never been carried out, i.  e. the so-called dredge mining. A feasibility study was prepared together with the Dutch firm IHC-MTI [10]. It showed that already today it would be possible at Lägerdorf to quarry the material up to a depth of 25 m below the lake level using standard dredgers. This underwater mining method was already included in the planning work for a future quarry site at Lägerdorf (Fig. 10).

The compact kiln concept has proved reliable in five plants resulting in operating experience of in part more than 20 years. Actually there is only the choice of modifying this concept for new projects. In the following the author will present his point of view for a reasonable strategy for streamlining projects in particular in Eastern Europe where most projects for chalk are waiting to be done in the years to come.

As regards the approach, first it has to be decided if the wet preparation of the raw material will be maintained (C, D, G) or one wants to or can give up the production of chalk slurry (E, F), respectively. Above all, the chloride load of the chalk has a decisive influence on the selection of the process.

Operating experience gained with LEPOL^© kilns at Lägerdorf has shown that the kiln operation is already impaired with chloride contents of more than 0.05  % (related to dry matter). In this case they already depend on the washout function of a filtration (D or G). The use of a chlorine bypass, which is possible in this case, would limit the future use of alternative fuels for the kiln right from the beginning and, in the opinion of the author, is less recommendable. Based on a taylor-made raw material preparation technology, the chloride window should be kept open as far as possible for the future use of alternative raw materials and fuels.

What are further reasons to choose the wet preparation for moist chalk, contrary to the general trends, and, consequently, the processes C, D, G?

1. Maintaining wet preparation in existing plants will considerably reduce the capital spending since the rehabilitation/extension will be carried out step by step.

2. The wet preparation of chalk, in contrast to the dry preparation, is more insensitive to deposit peculiarities (chloride load, separation of flints) and extreme climates (winter operation).

3. Wet preparation offers the greatest potential for future optimizations (technical and staff) in quarry operation.

4. Only the wet preparation of chalk offers the potential to achieve a minimum specific heat consumption (process G).

5. The process engineering risks are clearly lower with the processes C, D and G compared to E and F.

6. In addition to the provision of raw materials for cement making, the wet preparation of chalk makes it possible to additionally manufacture many different chalk products just from one preparation process, which could be exploited commercially.

If a decision is taken in favour of basically maintaining the wet preparation, the following approach in two steps would be a good choice:

1^st stage “Change”

– The entire quarrying and preparation system is only ­refurbished at first but in the end it is maintained as it is
for the time being. Capital spending will only be necessary
in case it is maintained as it is for the time being. Capital spending will only be necessary in case of an increase in performance.

– This first step with relatively low capital spending and risks makes it possible to commission and operate a modern kiln plant retaining as many proved processes as possible. At the end of this stage is the transition to a safe operation. A modern compact kiln including wet preparation is operated that replaces some old wet kilns.

– If the chalk deposit is chloride loaded (i.  e. on an average more than 0.05  % related to dry matter), a chlorine bypass has to be taken into account already in the 1^st stage “Change”. First it would serve as a valve for the amount of chlorine contained in raw material.

2^nd stage “Optimization”

– If a further reduction of the operating costs (e.  g. staff expenses in the quarry) is required, quarrying and preparation could be optimized efficiently.

– If the specific heat consumption for the clinker production is to be reduced to a possible minimum as regards the raw materials, there is the possibility of a transition to process G. The most exacting partial process from the process engineering point of view, i.e. the membrane filtration, for instance for chalk products, can be tested and levelled off a long time in advance.

– If there is already a chlorine bypass from the 1^st stage “Change”, now amounts of chlorine from alternative raw materials and fuels can be processed with its help.

From the point of view of the author, a transition to the dry process does not offer any process engineering advantages. Compared to the graded wet process, dry preparation of the
wet raw material chalk involves higher operating risks and
offers less possibilities of development. Finally, some remarks on the terminology of the clinker production processes are required.

The four basic processes of the production of cement clinker are usually defined by the kind and moisture of the kiln feed (Table 1). Though this definition ensures an immediate assignment in most of the cases, it does not apply to all processes (e.  g. E). A practicable definition must be open to new developments, i.e. must not orientate itself on existing, concrete technologies, but must base on processes and parameters. Actually, there are only two questions to be answered in the specified order (Fig. 11). Without any effort A/C can be assigned now to wet, B/D/G to semi-wet, E to semi-dry and F to dry processes.

In this connection it is necessary to remark on the system limits between the sub-processes raw material preparation and clinker burning in the overall process [11]. There must always be a decoupling material store between two sub-processes. This may be – as in the case of the classical wet process – a slurry reservoir, or – as in the case of the classical dry process – a raw meal silo.

For this reason the flash dryer of the compact kiln concept has to be assigned to the sub-process clinker burning. There is no material store between the flash dryer and the preheater. A container of 80 m³ between the flash dryer and the preheater in Chelm (process E) must not be seen as decoupling material store at this point but as preliminary proportioning container in front of the flow-regulating valve. If this definition is applied to the processes mentioned (A-G), Figure 11 will apply.

The terms „wet, semi-wet, semi-dry and dry process” reflect the development of the cement industry in the last 120 years as the transition from “wet preparation” to “dry preparation”. On superficial examination it is obvious to put the terms “dry” or “wet” on a level with “new” or “old”, respectively. If, however, a moist raw material such as chalk is dealt with, one should get rid of such conceptual prejudice.