Lamella
Lamella
Vol. IV
2004
No. 4
Nomenclature
, 0
0
pw
F, F1, F2
Q
q
d
dg
v, vg
f (d )
f (v)
(a)
1(a)
m,
(a)
(a, b)
d0, p, n
1. Introduction
Lamella tanks are now in widespread use in water and wastewater treatment installations. They belong to a group of sedimentation tanks whose running costs are
relatively low, and investment costs quite high. That refers particularly to conven-
W. P. KOWALSKI, R. MISO
tional rectangularly shaped or round Dorr clarifiers. When the running costs are on
a low level, it is possible to vastly reduce the investment costs.
That is why lamella tanks are receiving a great deal of attention from those sectors
of industry which, on account of process requirements or environmental considerations, handle large amounts of suspensions. This group includes the mining and minerals processing sectors and municipal utilities. The chief improvement in lamella
sedimentation tanks is an increased flow rate of suspension or enhanced sedimentation
efficiency, or the combination of both.
Three lamella sedimentation systems available presently are shown
schematically in Figure 3: counter-current, co-current and cross-current flow
systems.
The counter-current system, where the suspension flows in the direction opposite to
that of the sliding particles, is now most widely applied. Sedimentation proceeds in
conduits made of corrugated plates. A lamella packet of the envised type, manufactured in Poland, is shown in Figure 4. The lamella packet performance depends on the
relative length of conduits. Engineering the tubes with the relative length over 10 is
quite a difficult task.
The cross-current flow system comes next in the ranking list. In the cross-current
configuration, the suspension flows horizontally and the sediment flows along the inclined plates in the direction normal to that of the suspension movement. This arrangement seems to be most attractive because, unlike the counter-current systems, an
increase in the settling surface is not restricted by design data.
W. P. KOWALSKI, R. MISO
Figures 5 and 6 show counter-current packets used to modernise the existing rectangularly shaped tanks [9].
The parallel flow system, where the suspension flows downwards in the same direction as the settling particles, seems the least popular and its applications are but
a few because the clarified suspension and thickened sediment will mix while leaving
the sedimentation area. Nonetheless, a parallel flow system works really well as
a sludge thickener.
1. Rectangularly shaped tank with cross-current lamellas, consisting of a feed supply chamber, an overflow collector, an underflow collector or, alternatively, small
sediment collectors.
2. Support structure.
The feed supply chamber (Figure 7) 100 100 500 mm is provided with a perforated dosing tube (Figure 9) connected to a dosing valve (ball valve ) mounted in
a cover.
W. P. KOWALSKI, R. MISO
10
The valve is screwed indirectly in the chamber cover. The precise finish of the
cover edge ensures that the dosing tube coincides with the chamber axis. The holes in
the dosing tube open to the sidewalls and the wall opposite to the sedimentation
chamber. At the bottom the tube is secured with a cork which prevents the feed flow
into the supply chamber as well as mixing and drifting of sediment collected in the
first settling tray (Figure 9). Hence the mass of sediment retained in the supply chamber can be precisely established.
The supply chamber is separated from the sedimentation chamber by a removable
corrugated panel (Figure 12).
11
The corrugations on the dosing tube and the separating panel aimed at stabilising
the feed flow in the sedimentation chamber. The sedimentation chamber (Figure 5) is
equipped with a settling system and two suspended collecting hoppers with discharge
valves (ball valves ) (Figure 10).
On the skeleton structure hatches appear which seem to be an securing the supporting plates inclined at 55. Six such plates are mounted in that position (Figure 11).
Supporting plates are made of organic glass, 1.5 mm thick.
Sediment is collected by two hoppers dividing the tank bottom into two equal
parts. Hoppers have bolted ball valves . The design of the sedimentation chamber
allows mounting settling trays over the hoppers, these trays dividing the sedimentation
chamber into five equal settling zones. Sediment collected in trays can be subjected to
quantitative analysis as well as grain size distribution analysis. When no settling trays
are provided, the sedimentation chamber is equipped with an openwork insert which
can be fixed in the same position. The insert is made of organic glass rods of rectangular cross-section. The sedimentation chamber is separated from the overflow chamber by a thin panel wall made of organic glass, 5 mm thick.
Fig. 16. Experimental set-up: 1feed tank, 2mixer drive, 3outflow from the feed tank, 4peristaltic
pumps, 5conduits transporting the suspension, 6cross-current lamella packets, 7feed inflow to the
settling tank model, 9overflow collector, 10cross-current flow tank model
12
W. P. KOWALSKI, R. MISO
13
The solid phase content in the feed material ranges from 40 kg/m3 to over 120
kg/m3. The solid phase content in the overflow falls between 0.800 kg/m3 and 0.120
kg/m3. These results are consistent with our predictions. The effects of modernisation
can therefore be regarded as satisfactory. However, it is anticipated that solid phase
content in water treated in a Dorr clarifier might be exceeded in the future and that is
why the two-stage clarification is provided. The second-stage clarification proceeds in
a cross-current lamella tank.
2.3. Mathematical model of cross-current sedimentation
The starting point is Hazens theory of sedimentation and its generalisations suggested by Kowalski [20, 21], who has demonstrated that Hazens sedimentation theory
applies just as well to tanks with an inclined bottom [16] and that idealisation of suspension flow in elementary lamella conduits is responsible for slight undervaluing calculation results, at the same time the calculations become easier and less cumbersome.
Taking into account the specialists opinions [17, 26, 35], results of tests, calculations
and computer simulations, the authors provide below an algorithm based on the works
quoted above.
An assumption is made that the whole surface available in a sedimentation tank is
a major determinant of the process efficiency. When the tank is not filled with lamella
packets, the settling surface is taken as equal to the design value, in other words it
equals the surface of the water table. In a tank filled with lamella packets, two settling
surfaces are distinguished: that without and that with lamella packets (F1 and F2, respectively). The surface F2 contained in the packet conduits is obtained as the product
of the surface occupied by the packet layer and the specific surface factor ph. The specific surface factor ph is determined on the basis of design parameters of the lamella
packet and complex features of suspension flow (developing laminar flow, well developed laminar flow or flow with the rectangularly shaped velocity distribution pattern).
The specific surface factor ph indicates how many times the settling surface available
in a lamella packet is greater than the surface occupied by the packet. It was shown in
[17] that surface areas F1 and F2 are additive as long as certain assumptions are made.
For convenience the sum F1 + F2 is used in further calculations as the available settling
surface. Knowing the suspension flow rate Q, we obtain the surface load q equal to the
settling velocity of critical grains in the given process conditions:
q=
Q Q
= v ( d g ).
F1 + F2 F
(1)
The settling velocity v(dg) of critical grains is derived on the basis of the analysis of
particles flow in liquids (governed by Stokes formula), where and 0 stand for solid
phase and liquid phase density, respectively, 0 is dynamic viscosity and g is acceleration of gravity:
W. P. KOWALSKI, R. MISO
14
v (d g ) =
1 ( 0 ) g 2
dg .
0
18
(2)
Formula (2) yields the critical grain size dg. Knowing the type of statistical distribution of solid phase grain size f(d ) and the distribution parameters obtained from
grain-size measurements, we obtain the sedimentation efficiency :
dg
dg
= 1 f (d ) dd + d 2 f (d ) dd .
(3)
The actual formulation of (3) depends on the type of solid phase particle-size distribution f (d ). For some specific cases analytical solutions [22] are provided. When
the grain-size distribution f (d ) follows the log-normal pattern with the density function of the parameters m and (the mean value and standard deviation of the distribution of grain-size natural logarithms)
f ( d ; m, ) =
1 ln d m 2
1
exp
,
2
2 d
(4)
then for the argument a the analytical form of (3) related to the distribution function
N (a) holds:
ln d g m
(d g ; m, ) = 1 N
ln d g m
+ exp{2 [ 2 (ln d g m)]} N
2 .
(5)
When the solid phase particle-size distribution f(d) follows the generalised gamma
distribution with the density function of the parameters d0, n, p (scale parameter and
shape parameters):
n
f ( d ; d 0 , p, n ) =
d 0 ( p)
d
d0
pn 1
d n
exp ,
d 0
(6)
d n
2 dg
g
p,
,
2 p+
n d 0
d 0 d
+ 0
( d g ; d 0 , p, n ) = 1
dg
( p)
( p)
15
(7)
(a) is a well-known Eulers gamma function and (a, b) is the incomplete Eulers
gamma function:
b
( a, b) =
1
x a 1 e t dt.
(a) 0
(8)
d
b = g
d0
(9)
In order to compute the sedimentation efficiency from (5) or (7), we ought to know
the parameters of the grain-size distribution and the critical grain size dg. Parameters
of the statistical distribution of the solid phase particle size are obtained from granulometric analysis [18, 19].
Theoretical considerations of computations of sedimentation efficiency of polydispersive suspensions seem rather complicated though the available computer techniques render the task feasible. All the same, the model can be simplified by assuming
one of the boundary velocity distributions, so as to have it developed laminarly or
rectangularly shaped. The calculation procedure becomes less complicated, though
less accurate. According to the authors, this inaccuracy is fully acceptable in the investigations of dilute suspensions.
W. P. KOWALSKI, R. MISO
16
= f (q, m, , d g ),
(10)
where
Q
= v (d g ),
F
(11)
q=
dg =
18 0
q.
( 0 ) g
(12)
0 =
1.79 10 3
,
1 + 0.0337 + 0.000222 t 2
(13)
where t stands for suspension temperature [C]. The suspension temperature is assumed t = 20 C and the dynamic viscosity 0 is equal to 110 3 kg/m/s.
Selected results of computer simulations are shown in Figures 1926. Sedimentation efficiency is plotted as the function of surface loading (in the range of 01.5
m3/m2/h) for the specified values of log-normal parameters: m = 2.0, 2.5, 3.0, 3.5 and
= 0.6, 1.0. In each case, five curves are plotted to show the influence of the solid
phase particles density = 2000, 3000, 4000, 5000, 6000 kg/m3.
It is readily apparent that the influence of the surface loading is essential. In the
case of suspensions containing the finest solid particles (m = 2, = 0.6, see Figure 17)
the sedimentation efficiency over 0.80.9 is achievable, provided that surface loading
17
cannot exceed the value of 0.15 m3/m2/h, no matter what the solid phase density. The
influence of the solid phase density is vital when the sedimentation efficiencies
achieved turn out to be low. For example, for a surface loading of 1 m3/m2/h the sedimentation efficiency of particles with the density of 6000 kg/m3 is 0.6, while for the
particles density of 2000 kg/m3 it will be slightly more than 0.2.
18
W. P. KOWALSKI, R. MISO
19
stage treatment process in a Dorr clarifier. The input data for the design are the results
of investigations and simulations summarised in the previous chapters.
Suspension flowing at the rate of approximately 120 m3/h enters the cross-current
tank, about 20 m3/h of the suspension will be discharged via an underflow. Thus we
obtain the flow rate of about 100 m3/h of the suspension with the solid phase content
nearing zero.
Extensive tests and simulations reveal that surface loading in the tank should not
exceed 0.25 m3/m2/h and hence the settling surface should be at least 400 m2. It is suggested that two twin tanks be built. Application of 25 lamella packet segments is
feasible, thereby increasing the settling surface 25-fold in relation to the surface occupied by the packets. Let the base of the working unit of one tank have the surface area
of 10 m2. The available settling surface will be 250 m2, leaving a safety margin of
20%.
In the preliminary stage of design, it is assumed that the tank housing will be made
of steel, though other options are considered, too. The housing might be also made of
plastic materials and that solution offers several benefits: resistance to corrosion and
20
W. P. KOWALSKI, R. MISO
vastly reduced maintenance costs. However, the cost of constructing a plastic frame
will be decidedly higher.
The computer model of a cross-current sedimentation tank and the selected details
are shown in Figures 2731. The computer model is shown as a single still from
a computer-processed film showing the spatial view and tank interior as well as subsequent steps of the tank assembly.
21
22
W. P. KOWALSKI, R. MISO
chamber. The flow path equals 1600 mm (half-width of the chamber). From the
chamber filled with lamella packets the suspension flows to the overflow basin.
13. Overflow basins are fixed along the longer walls of the split chamber. On one
of the shorter walls there is the trough connecting the overflow basins so that the
whole clarified suspension leaves the system.
14. Each of the twin chambers is divided into 10 sections by vertical panels made
from corrugated plastic plates, with the wave height of 50 mm. Dividing panels are
supported on shelves on the frame level and extend over the liquid table (over 100
mm). Single section dimensions are: 1600 mm (split chamber width) 500 mm (section width taking into account the width of a corrugated panel).
15. Each section (including the first and the last one) is surrounded by two parallel
corrugated baffles. One baffle is used between each two neighbouring sections (apart
from the first and the last one).
16. At the bottom of each section an extra bottom surface is provided: a protective
barrier directing the suspension towards the section interior. The barrier is made of
PVC board shaped like a sloped roof inclined at 60o with a horizontal ridge 40 mm in
width. The barrier length equals the width of the split chamber (identical to the section
length equal to 1600 mm). The barrier is supported on shelves lying on the frame.
17. Each section has two saw-tooth edges, their height variable and controllable
over 50 mm to ensure the equal chamber loading.
18. On the end of feed basin each section is provided with a guiding barrier,
directing the suspension downwards. The guiding barrier is located at the distance of
100 mm from the feed supply basin and extends over the liquid level (about 100 mm),
the remaining part (i.e. 500 mm) should be immersed.
19. On the overflow basin end each section has a guiding barrier different from
the previous one in that that it should be immersed to the depth of about 200 mm.
20. In each section, there are lamella packets. Packets are arranged in a herringbone pattern, the elementary unit being a symmetrically shaped profile resembling
a roof inclined on both sides and with a flat ridge. The inclination angle of the profile
is 60o. The terminal edges are bent at the right angle so that each profile be supported
on the previous one. In the bend sections, some openings are cut, accounting for 80
90% of the bend surface. The single packet contains 25 shaped profiles. Shaped profiles are interconnected by two vertical tubes 1500 mm in length and with the outside
diameter 10 mm. These tubes pass through the openings made in the ridges of the
shaped profiles. Between the subsequent profiles there are spacing elements (bushings) with washers. The height of the spacing elements is 40 mm. In the lowermost
tube section, there is a back nut with a washer. In the topmost tube section, a hook is
provided to hang the packet. Hooks are hung on a rod coinciding with the section axis,
at the height of 200 mm over the liquid level. A screw joint between the hook and the
tube enables position control of the packet inside the section (in a range of about 50
mm). The rod is supported on a framework connected to the split chamber. Inside the
section the packet is tangent to the ridges of corrugated baffles.
23
4. Conclusions
Theoretical studies and computer simulations of cross-current processes in sedimentation tanks with lamella packets reveal that tank capacity during clarification of
dilute suspensions can be vastly improved if compared with the conventional tanks
(even ten-fold increase of process capacity is reported). As regards the lamella
counter-current processes, the improvement in performance is two- or even three-fold,
in relation to the space they occupy.
It appears, therefore, that cross-current sedimentation tanks can successfully be applied in industrial installations for clarification of suspensions which, on account of
the presence of very fine particles or the flow rates, are hard to process because of
space limitations and high costs. In many sectors of industry, suspensions are clarified
in installations designed many years ago and the achievable sedimentation efficiencies
are regarded as insufficient in the light of the present standards. In such cases, the use
of cross-current lamella tanks as secondary clarifiers might be the answer to the problem, ensuring the required quality of clarified suspension at relatively low costs, estimated to be 20% of costs involved in construction of traditional tanks.
High capacity of cross-current lamella tanks (or improved sedimentation efficiency) are achievable by extending the settling surface thanks to the placing of lamella packet sections at small intervals. The presence of lamella packets causes no
problems during the clarification of dilute suspensions though when thickened suspensions are handled, the clarification processes might be disturbed.
References
[1] Bandrowski J., Hehlmann J., Merta H., Zioo J.: Opracowanie metody optymalnego doboru wysokosprawnych osadnikw z wypenieniem na podstawie dowiadczalnych bada
osadnikw, In. Ap. Chem.,1 (1997), 38.
[2] Bandrowski J., Hehlmann J., Merta H., Zioo J.: Podzia zawiesin ze wzgldu na moliwo zagszczania w osadnikach z wypenieniem oraz analiza czynnikw wpywajcych
na proces sedymentacji cienkowarstwowej, In. Ap. Chem., 3 (1997), 37.
[3] Bandrowski J., Hehlmann J., Merta H., Zioo J.: Studies of sedimentation settlers with
packing, Chem. Engng. and Proc., 36, (1997), 219229.
[4] Bandrowski J., Merta H., Zioo J.: Analiza modeli osadnikw z wypenieniem, In. Ap.
Chem., 6, 1995.
[5] Bandrowski J., Merta H., Zioo J.: Sedymentacja zawiesin. Zasady i projektowanie, Wyd.
Politechniki lskiej, Gliwice, 2001.
[6] Bandrowski J., Merta H., Zioo J.: Analiza modeli osadnikw z wypenieniem (in Polish),
In. Ap. Chem., 6, 1995.
[7] Bandrowski J., Merta H., Zioo J.: Modele osadnikw z wypenieniem, Mat. XVI
Oglnopolskiej Konferencji Naukowej Inynierii Chemicznej i Procesowej, t. 1. Przepywy, mieszanie. Procesy dynamiczne z faz sta, s. 180185, Wyd. Politechniki
Gdaskiej, 1215 wrzenia 1995.
24
W. P. KOWALSKI, R. MISO
[8] Buer T., Margraf M.: Enhancement of activated sludge plants by lamellas in aeration
tanks and secondary clarifiers, Paper presented at the Aeration Conference on Applied
Techniques to Optimize Nutrient Removal and Aeration Efficiency, Helsinki, 2000.
[9] Boycott A.E.: Sedimentation of Blood Corpuscles, Nature, 104, 532.
[10] Camp T.R.: Sedimentation and the Design of Settling Tanks, Trans. Amer. Soc. Civ.
Engrs, 111, 1946, 895958.
[11] Gga J.: Wysokosprawne osadniki z wkadami rurowymi do oczyszczania ciekw hutniczych, Zesz. Nauk. AGH, 561, 9, 1976, 8999.
[12] Haba J., Nosowicz J., Pasiski A.: Klren und Eindicken von Suspensionen in Lamelleneneidickern, Aufbereitungs-Technik, 4, 1980, 198201.
[13] Haba J., Nosowicz J., Pasiski A.: Sedymentacja zawiesin w osadnikach pytowych, Rudy
Metali, 23, 9, 1978, 440444.
[14] Haba J., Nosowicz J., Pasiski A.: Wyznaczanie stopnia sedymentacji w osadniku z
wypenieniem pytowym, Rap. Inst. In. Chem. Urz. Ciepl. Polit. Wroc., seria preprints,
91, Wrocaw, 1980.
[15] Haba J., Pasiski A.: Badanie osadnika pytowego w zastosowaniu do technologii
otrzymywania wodorotlenku glinu, Rap. Inst. In. Chem. Urz. Ciepl. Polit. Wroc., 4,
Wrocaw, 1979.
[16] Hazen A.: On Sedimentation, Trans. ASCE, 53, 1904, 4588.
[17] Kowalski W.: Mathematical model of sedimentation process in the bended suspension
stream (in Polish), Arch Ochr. rod., 12, 109120, 1991.
[18] Kowalski W.: Parameters of the particles composition of the emitted dust from iron
metallurgy processes (in Polish), Arch. Ochr. rod., 12, 6779, 1991.
[19] Kowalski W.: Sedimentation balance interpretation method of information about particle composition (in Polish), Arch. Ochr. rod., 12, 1991, 159167.
[20] Kowalski W.: Generalization of Hazens Sedimentation Theory, Archives of Hydroengineering, Vol. XXXIX, 2, 1992, 85103.
[21] Kowalski W.P.: Analiza teoretyczna i badania procesu sedymentacji wielostrumieniowej
(in Polish), Problemy Inynierii Mechanicznej i Robotyki, 3, Krakw, 2000, 179.
[22] Kowalski W.P.: Podstawy teoretyczne projektowania osadnikw z wkadami
wielostrumieniowymi (in Polish), Zeszyty Naukowe AGH, seria Mechanika, 27, Krakw,
1992, 132.
[23] Kowalski W. P.: Modelowanie matematyczne procesu sedymentacji ziaren zawiesin
polidyspersyjnych, Mat. XLIII Sympozjonu pt. Modelowanie w mechanice (in Polish),
Gliwice, 2004.
[24] Kujawska E.: Badania procesu sedymentacji w osadniku z wypenieniem pytowym
i profilowym, PhD Thesis, Politechnika lska, 2003.
[25] Niedwiedzki Z.: Badania teoretyczne i eksperymentalne wypenie osadnikw wielostrumieniowych, Zesz. Nauk. Politechniki dzkiej, 863, d, 2000, 162.
[26] Merta H.: Projektowanie osadnikw z wypenieniem. Cz 1. Parametry projektowe
i technologiczne osadnikw pytowych i rurowych oraz ich klasyfikacja (in Polish), In.
Ap. Chem., 2, 1996, 310.
[27] Nipl R.: O zastosowaniu uoglnionego rozkadu gamma do aproksymacji krzywych
skadu ziarnowego, Mat. XIII Krakowskiej Konferencji Nauk.-Techn. Przerbki Kopalin
(in Polish), Krakw, 1979, 323330.
25
[28] Olszewski W.: Osadniki wielostrumieniowe, [in:] Nowa Technika w Inynierii Sanitarnej
(in Polish), Wodocigi i Kanalizacja, t. 5, 1975.
[29] Oden S.: Eine neue Methode zur Bestimmung der Kornerverteilung in Suspensionen, Kolloid -Z , 18, 2, 1916, 3347.
[30] Orzechowski Z.: Przepywy dwufazowe, jednowymiarowe, ustalone, adiabatyczne, PWN,
Warszawa, 1990.
[31] Papoulis A.: Prawdopodobiestwo, zmienne losowe i procesy stochastyczne, wyd. l,
WNT, Warszawa, 1972.
[32] Miso R.: Badania prostopadoprdowego procesu sedymentacji, PhD Thesis, AGH,
Krakw, 2000.
[33] Stacy E.W.: A Generalization of the Gamma Distribution, Annals of the Mathem. Statistics, Vol. 33, 3, 1962.
[34] Stokes G.G.: On the Effect of the Internal Friction of Fluids on the Motion of Pendulums,
Camb. Trans., Vol. 9.
[35] Materiay informacyjne firmy Sala Inc.
[36] Zioo J.: Influence of the system geometry on the sedimentation effectiveness of lamella
settlers, Chem. Engng. Sci., Vol. 51, 1, 1996, 149153.
Prostopadoprdowe osadniki wielostrumieniowe
Przedstawiono genez urzdze sedymentacyjnych opartych na wykorzystaniu efektu Boycotta. Opisano trzy podstawowe ukady, w jakich prowadzi si sedymentacj wielostrumieniow. S to: ukad przeciwprdowy, prostopadoprdowy i wspprdowy. Popraw efektywnoci sedymentacji w osadnikach z wkadami wielostrumieniowymi osiga si, zwikszajc
powierzchni sedymentacyjn.
Miar zwikszenia powierzchni sedymentacyjnej jest wskanik wyraony przez stosunek
pola powierzchni sedymentacyjnej zawartej w pakiecie do podstawy pakietu, nazywany powierzchni waciw. Przedstawiono typow konstrukcj pakietu wielostrumieniowego dla sedymentacji przeciwprdowej. Ograniczenia konstrukcyjne (zwikszanie dugoci przewodu
z jednoczesnym zmniejszaniem jego przekroju poprzecznego) i wykonawcze powoduj, e nie
da si osign wskanika powierzchni waciwej o wartoci wyszej od 56, a ponadto przepyw zawiesiny w przewodach o dugoci wzgldnej l/d > 10 osiga zakres duych liczb Reynoldsa i przestaje by laminarny. Mona jednak zastosowa inn koncepcj rozwizania konstrukcyjnego pakietw opart na wielostrumieniowej sedymentacji prostopadoprdowej.
W tym przypadku ograniczenia konstrukcyjne i wykonawcze nie stanowi przeszkody, pakiet
jest, bowiem stosem pyt o teoretycznie nieograniczonej wysokoci.
W artykule opisano oryginalne stanowisko do bada sedymentacji prostopadopradowej.
Przedstawiono take matematyczny model procesu sedymentacji prostopadoprdowej, wyniki
bada oraz wyniki symulacji komputerowych. Wyniki bada i symulacji komputerowych prowadz do wniosku, e w osadnikach tego typu mona nawet ponad 10-krotnie zwikszy wydajno lub adekwatnie podnie efektywno sedymentacji. Na podstawie bada i oblicze
przedstawiono komputerowy projekt osadnika prostopadoprdowego o wydajnoci 100 m3/h.
Urzdzenie tego typu zajmuje okoo 10% miejsca niezbdnego dla tradycyjnych urzdze sedymentacyjnych i pozwala osign du efektywno porwnywaln z efektywnoci tradycyjnych urzdze. Dlatego moe by stosowane do wysokosprawnego i wysokowydajnego kla-
26
W. P. KOWALSKI, R. MISO