“HOLDERBANK” Cement Seminar Process Technology Homogenizing SilosSUMMARY Homogenizing silos are classified according to the mode of the applied working principles, in - batchwise mixing silos - continuous overflow silos - continuous mixing silos. Batchwise mixing silos show a better homogenizing efficiency than con- tinuous mixing silos. The disadvantage is the higher energy consumption and above all the high investment costs. The homogenizing efficiency of continuous working silos is a function of the kind of disturbances. Periodical long-term fluctuations (75 h) cannot be reduced sufficiently by the silo. This situation can only be improved by cutting the long-term Fluctuation, i.e. by adjusting the raw mix proportioning system. Sudden peaks can be damped most efficiently by an immediate counter-action, i.e. by producing a contrary peak.TABLE OF: CONTENT — ws 1.2 1.3 24 2.2 2.3 TYPES OF HONOGENIZING SILOS Discontinuous batch-homogenizing silos a) The Polysius octant system b) Other aeration systems c) The FLS-funnel flow system d) Technical specifications ) Homogenizing efficiency Continuous overflow homogenizing systems Continuous homogenizing silos a) Claudius Peters mixing chamber silo b) Claudius Peters homogenizing chamber silo c) Polysius continuous homogenizing silo d) Polysius mltiflow silo e) IBAU central chamber silo f) FLS controlled flow system g) Partec continuous blending silo OPERATION OF CONTINUOUS HOMOGENIZING SILOS The silo investigations Results of the silo investigations The silo operation BIBLIOGRAPHY DIAGRAM 1, 2, 3 PAGE 10 12 13 14 15 16 7 18 20(a TYPES OF HOMOGENIZING SILOS Homogenizing silos are classified according to the kind of the working principles, in = batchwise mixing silos - continuous overflow silos = continuous mixing silos. In the following some of the most applied silos of the three types are discussed. The technical data represent approximate values. Discontinuous batch-homogenizing silos A typical arrangement of silos consists of two raw meal storage and two homogenizing silos. The top silos are the homogenizing silos and the lower ones are storage silos. The homogenizing silos are sized for approximately 6 to 11 times the raw mill capacity. The silos are intermittently filled and during the time one silo is filled the ma- terial in the other silo is homogenized. Emptying one silo takes approximately two to. three hours. Nee , q in a 4 d i ry | RET |Most commonly used are quadrant or octant aeration systems. The following figure shows as an example the basic principle of a Polysius octant system. material input aerated Ga non aerated foThe silo bottom is divided into 8 sectors of which only two oppo- site are aerated at the same time, i.e. only one quarter of the total silo area is aerated at the same time. Moreover, the sectors are divided’ into several segments (a, b, c, d), where different air mass flows can be introduced. The aeration time of one sector is about 10 minutes and the whole homogenizing time 40 minutes. b)__ Other aeration systems Other batch homogenizing systems work more or less similar to the Polysius octant system. Two of them are presented in the following: Fuller quadrant method active Passive The bottom of the silo is divided into four quadrants. Three compres- sors are installed. Two of them blow into the one active while the remaining one blows into the other three passive quadrants. After a set time periode they are switched over to the next quadrant series. Moller shearing flow method active In principle the same as the Fuller method is the so-called shearing flow method of Méller with a different distribution of the sectors.A double deck system without aeration for homogenizing purposes is the funnel flow silo of FLS. In the filling phase of the silo the raw meal is stored in layers. During the outlet of the material a funnel is built in which material of different inlet layers slides. Thereby a certain mixing effect takes place. JN e The homogenizing efficiency is dependent on the angle o of the raw meal in the silo. The steeper this angle the higher the mixing effect. Due to the fluidification of the raw meal during transport this angle @ becomes, however, very flat. Therefore at least two silos in series should be provided so that the mixing effect is doubled. pect Batch-type homogenizing silos have been installed up to capacities of approximately 5000 t. The diameter to height ratio (effective height of material inside silo) is usually in the order of 1:1,2. S Power consumption obtained in batch-type systems depends to a large extent on homogenizing time required and detail design of the aeration systems. With the octant system values as low as 0.3 kWh/t of raw meal have been obtained. Generally, the power consumption reached for batch homogenizing varies in the range of 0.3 to 0.6 kWh/t and it should be mentioned that the lower values are usually obtained with the octant system.e)__.Homogenizing efficiency The homogenizing efficiency, defined as the ratio between the stand- ard deviation of the chemical concentration of the inlet and of the outlet, is with batch systems excellent and values up to 1:15 have been obtained in practice. The problem of batch homogenizing silos is usually not the homogeni- zing efficiency within one batch, but the change in the mean chemical concentration from one batch to the other. This change is avoided if, in the average, during the filling time the desired mean value is achieved by corresponding adjustments of the raw mix ratio. concentration inlet time change between batches concentration outlet1.2 Continuous overflow homogenizing system The overflow system usually consists of one homogenizing silo in combination with a raw meal storage silo. The overflow silos are usually sized for approximately 6 to 10 times the raw mill capacity and the diameter to height ratio is similar to the batch-type homogenizing silos, in the order of 1:1.2. The following figure shows the basic principle of an overflow homo- SECTION B-B genizing silo. dividing wall bottom segments with strong and weak aeration ‘fs material flow By different aerated bottom segments the raw meal is mixed and at the same time transported from the inlet part to the other side of the dividing wall, where the outlet of meal is located. The power consumption of the overflow system is normally higher than of the discontinuous batch-type systems and values of approximately 1.2 kih/t have been reached.1.3 Continuous homogenizing silos The modern design of continuous homogenizing systems combines raw meal storage and homogenizing in one silo. Separate silos, as re- quired for the batch or overflow system are not necessary. Instead of homogenizing the material pneumatically as described before, blending occurs during the time when the material is discharged from the silo by creating a funnel. funnel A sufficiant homogenizing effect can therefore only be obtained if a certain filling height in the silo is maintained. At low filling degrees the homogenizing effect is rather low and it is in practice therefore required to keep the silo filling degree above at least 30 8. Various systems have been developed in the meantime. These are: = Claudius Peters mixing chamber silo - Claudius Peters homogenizing chamber silo = Polysius continuous homogenizing - Polysius multiflow system = FLS controlled flow system - IBAU central chamber silo = PARTEC continuous blending silo.a)___Claudius Peters mixing chamber silo The mixing chamber silo has been applied in several cement plants, mainly in Europe. The following figure shows the basic principle of this silo. aerated sector quadrant aeration system As can be seen, the material entering the silo is evenly distributed thus forming horizontal layers of material. The silo bottom is divided into 12 sectors equipped with air pads and a mixing chamber with a quadrant aeration system. Air supply to the one active outer sector and one active quadrant alternates in rotation and has been synchro- nized. For discharging material from the silo, one outer sector and one active quadrant of the mixing chamber are aerated at the same time, thereby creating a funnel above the aerated sector. The homogenizing effect is reached by = sectorwise formation of funnels caused by gravity discharge so that different horizontal layers slump into each other, - generating a turbulent flow of material in the mixing chamber due to pressure relief and rather intensive aeration. Mixing chamber silos have been installed up to capacities of approxi- mately 10'000 t and there are several silos in operation. Diameter to height ratio is usually in the range of 1:1,2 to 1:1,5.The specific power consumption obtained is in the order of 0.3 klih/t or lower. The homogenizing efficiency usually obtained with one silo is in the range of 1:2 to 1:5 (silo to be filled with at least 30 % of material). Homo ensing- chamber slo: plan- 10 - Comparable to the mixing chamber silo a better homogenizing effici- ency is expected by - enlargement of the outer ring for more efficient formation of mixing funnels, - application of a relatively large homogenizing chamber with im- proved aeration system, - secondary material fed through the roof of the homogenizing chamber. So far no practical experience is available from this new silo type, particularly concerning the blending efficiency. Claudius Peters expects the efficiency to be substantially better than for the mixing chamber silo. The power consumption expected is in the range of 0.6 khih/t. Polysius continuous homogenizing silo The figure below shows the basic design of the Polysius system. aad homogenizing fillingThe special design of the Polysius homogenizing silo is shown below: Legend Feeding system Continuous homogenizing silo 1 2 3 Discharge channel 4 Openings to the discharge channel Cover of discharge Channel Aerated areas with one sided ‘inclination 8 Transport to kiln The silo bottom consists af sloped and aerated channels and a relatively small central chamber. Number of channels depend on silo diameter. There are approximately three openings per channel, the remaining portion of the channels being covered. The largest open- ings are at the silo wall. The surfaces ‘between the channels are sloping and partly covered with aeration pads. The inlets of the meal are 90°C dephased from the aerated sectors. Material is discharged from the silo by aerating one or two diametri- cal channels, and therefore creating funnels above the aerated discharge openings. The homogenizing effect is reached by - formation of multifunnels caused by gravity discharge - alternating channel discharge in rotation. Polysius continuous homogenizing silos have been installed up to capacities of 7500 t and there are several silos in operation. Diameter to height ratio is usually in the range of 1-1,5. The specific power consumption obtained is in the order of 0,2 kWh/t. Homogenizing efficiency obtained with one silo is in the range of 1:2 to 1:5 (silo to be filled with at least 30 % of material). Extraction and dosing equipmentqa. The multiflow system is the improved continuous homogenizing system from Polysius with basically the following modifications: - The horizontal cover above the channels has been altered to a sloped cover parallel to the inclined channels, - the central chamber bottom was lowered, - the sloped surfaces between the radially arranged channels have been omitted. With the above mentioned modifications, a more extensive activation of the outer ring zones is expected. Legend 1 Silo volume 2 Central chamber 3 Discharge openings 4 Coverings 5 Openings to discharge channel 6 Aerated discharge channel 7 Distributing system of feede) The homogenizing effect is similar to the other silos created by sectionwise formation of funnels caused by gravity discharge. IBAU_central chamber silo wt 7 2 FUNNEL + ; + 3 ‘BELUFTUNGSSEKTIONEN | | 4 ll FLOW-CONTROL-GATE I 2 ENTSTAUBEA eee 4a} | + + at lie / i= | Mh 4 . _ The material entering the silo is evenly distributed and the bottom sections are alternately aerated for sectional discharge of material. The power consumption is about 0.2 kWh/t. The homogenizing efficiency ‘is in the same order as previously mentioned (appr. 1:5).£)___FLS_controlled_flow system The controlled flow system has been based on the idea that the entire material colum inside the silo must be kept moving and a suitable retention time distribution of the material entering and leaving the silo must be reached in order to obtain a good homogenizing effect. The figure below shows the design of the silo. Material entering the silo is evenly distributed. The silo bottom consits of 42 triangular segments and seven discharge openings. The triangular segments are partly covered with aeration pads which can be independently aerated. From the seven outlet openings material can be discharged at a controlled flow rate and fed to a pneumatically aerated mixing vessel placed under the silo.The technical data for this silo, given by FLS are 0.3 kih/t and a homogenizing efficiency of 1:10 for a silo filling degree above 50 %. This value is not yet proved in practice since the system has not yet been installed so far. 9... The Partec system has been developed in Finland and a piggyback silo system consisting of two 3000 t silos is in operation. The figure be- ow shows the basic principle of this system. Sectional discharge from the silo is by special design of rotary feeders over the radius of the silo. construction of the silo bottom ridges with aeration bearing drivin moto rotary feeder air slide Due to this special discharge system a good homogenizing efficiency can be expected. Mechanical problems are anticipated with large dia~ meters.OPERATION OF CONTINUOUS HOMOGENIZING SILOS The blending efficiency of a homogenizing silo is given by the ratio between the standard deviation of the chemical fluctuations: on the inlet sq and of the outlet sw. Sa Sw where s= ——» concentration of the single sample No i mean value 3 xix —= number of samples The diagram 1 shows as an example the fluctuations of the lime sa- turation in function of the time. In the first part of the observed interval the inlet and the outlet fluctuation have the following values: standard deviation s This efficiency does not mean that all peaks coming into the silo are reduced by the factor 8.8. The second part of the observed inter- val illustrates such a case. A very great peak with a maximum lime saturation value of 112.0 was produced and entered the silo. The corresponding concentration at the outlet rises thereby up to 103.5. the blending efficiency of the silo, calculated with the two maximum values is 112.0 - 96.0 403.5 -96.0 en =2.124 The suddenly appeared peak is much worse homogenized than the more or less stochastic fluctuation of the first part of the ob- served interval. This example shows that different kinds of fluctuations are homo- genized with different efficiencies. The blending efficiency de- fined by the inlet and outlet standard deviation is, thus, only a very general indication. To know the specific behaviour, it is required to investigate the proceedings in the silo. For this rea~ sons extensive tests were made by TC/VA, which are summarized in the following. The silo investigations The investigations of the proceedings in a silo are not very easy be- cause a direct observation of the flow during the operation is impos- sible. The only way was to use the information from the inlet and out- let concentration. For this purpose a well defined disturbance of one of the entering components was produced by changing its raw mix ratio. Thereby a stable and constant component had to be taken, which is only in little proportion in the mix and which has small fluctuations, for example the iron oxide. Now the concentration at the silo-outlet was compared with the disturbed inlet. The question is now, which kind of inlet disturbance is most represen- tative for the silo? For this aim it is important to know what kind of disturbance can enter the silo. There are two basic cases: - suddenly appearing disturbances - periodic oscillations. a)___Suddenly appearing disturbance single impulse function ¢ S & € o 8 2 5 8 time2.2 A suddenly appearing disturbance can be divided into a sum of single impulse functions. The knowledge of the silo answer (concentration outlet) to one of such inlet impulse function allows to predict the silo outlet of any suddenly appearing disturbance. If a component of the entering meal behaves like a harmonic oscilla~ tion, for example a sinus function, with the amplitude A, and the frequency w, the exit function will also be harmonic with the same Frequency « but with an other amplitude A, and with a phase shift 'p . inlet outlet concentration ime In reality a periodical disturbance is not exactly a harmonic func- tion but according to the law of Fourier each periodical function can be divided into a sum of harmonic functions. The silo answer to a harmonic silo inlet function can be calculated from the answer to a single impulse function. (Fourier- or Laplacee @ Transform). Therefore, with an impulse function as inlet disturbance, the behavi- our of the silo can be sufficiently investigated and the blending efficiency for all cases predicted. Results of the silo investigations The described test with the impulse function as inlet disturbance was made on a Claudius Peters silo. With the results a mathematical model has been formed. In the following the model of second order is explai- ned.The incoming raw meal builds horizontal layers on the surface. When a sector on the bottom is aerated, on the surface over this sector a fun- nel is built and from the different surface layers the raw meal slides ‘into this funnel. Thereby the first and most significant mixing effect 2\\ [Z After a certain time the aeration switches over to an other sector and a new surface funnel in another aera is built. Afterwards the raw meal is transported down to the mix chamber. This transport takes place in a "tube" with a very small diameter. funnel transport mix chamber In the tube nearly no cross mixture with the other material takes place and, thus, this part has an inferior influence to the homogenizing effi- ciency. In the mix chamber the raw meal is homogenized again. The volume of this chamber is rather small, so that the retention time of this part is only 0.4 Ch]. The behaviour of the continuous homogenizing silo can be described by a so-called linear model of second order allowing to express the homoge- nizing efficiency by the following equation:2.3 e= where Ty = 4.1 Oh] T, = 0.4 [hI The constant T, and T, represent the average retention time of the raw meal in thé main Romogenizing parts of the silo. The most impor- tant influence is given by the constant T, describing the proceedings on or near the surface of the silo. It is clearly evident from this equation that the homogenizing effi- ciency of the silo is a function of the frequency. of the ingoing fluctuations. In the diagram 2 the mathematical model of second order is presented. A comparison with a model of first order neglecting the influence of the mix chamber shows that this element has only an influence on short- term disturbances. The silo operation The diagram 2 shown that all fluctuations with a periodical time shor- ter than about 5 [h] are homogenized better than 1:5. The first part of the iTlustrated fluctuation of diagram 1 is exactly such a short @y time fluctuation. The homogenizing efficiency is therefore good and a value over 1:5 can be expected. Longtime fluctuations with periodical times over 5 [h] are homogenized unsufficiently. For example an oscilla- tion of the mean value with a periodical time of 20 [h] is only homo- genized with 1:1.6. concentration timeconcentration c The figure above shows the inlet and outlet function of this case. The longtime oscillation of the mean value is damped by the ratio 1:1.6 meanwhile the overlapped shorttime fluctuations are homogenized with a much better ratio. In this case the silo is not able to sufficiently homogenize the raw meal. The only way to improve the situation is to cut the long-term fluctuations by corresponding adjustments of the weigh feeders of the raw mix porportioning. The adjustment frequency should be choosen as short as possible so that only short-term disturbances are produced. Different from these disturbances with more or less clear defined fre- quencies are sudden peaks (see diagram 1). These peaks are produced by a lot of reasons like > new cut of a mix bed - stop of one of the material components - unadequate adjustment of the feeders by the operator etc. By the assumption that the stT@:works as“a mddeT“of finstsorder ‘the -out- Jetfunction-to:an. inlet peak can be determined: time t time t Diagram 3 shows the effective and the calculated outlet function. The conformity is in spite of the rough assumptions sufficient. In this case the silo is not able to homogenize the inlet peak suffi- cently. Therefore, counter-actions must be taken by making adequate adjustments of the raw mix ratio. The sooner the reaction takes place the more the inlet peak is reduced. Furthermore, only a reciprocal cor- rection, i.e. a correction with a contrary peak is able to effectively damp the peak. If instead only a normal correction towards the desired mean value is made the peak goes nearly undamped through the silo.intet peak concentrgtion contrary peak time time The above diagram shows the adequate correction of a peak by a corresponding contrary peak. To summarize: 1) Long-term fluctuations ( 5h) are not reduced sufficiently by the silo and have to be evened by adequate adjustments of the raw mix ratio. 2) Sudden peaks are damped by a sudden counter-action producing on purpose a contrary peak.Res ine reg Monee (P32 Asya rary > E 2 peery sy ve 2 eeeoe Fo re % . mee Loe rf z meebo sor z meet os vy 6 z tBIBLIOGRAPHY (1) [2] [3] R. Hasler / J. Waltisberg Mischwirkung von Durchlaufhomogenisiersilos: Untersuchung am Mischkammersilo in Hover (VA 79/4677/) J. Waltisberg Results of the investigation on the homogenizing efficiency of continuous blending silos (VA 80/4707/E) J. Maltisberg Mathematisches Modell des Mischverhaltens eines kontinuierlich arbeitenden Homogenisiersilos, entwickelt am Beispiel des CPAG- Silos in Hover (VA 80/4718/D)