How to divide a Pneumatic Conveying Stream

Pneumatic Conveying Lines

How to divide a Pneumatic Conveying Stream

Splitter for continuous distribution of bulk solid material during pneumatic conveyance - dimensioning, calculation, operating behaviour. Worldwide, a great number of so-called splitters are in operation as elements of pneumatic conveying lines, in order to continuously distribute the conveyed bulk material to several receiving stations. These splitters have taken an important position within the pneumatic conveying systems and are applied in different designs. But there is not much literature available dealing with the phenomenon “splitter”, and providing assistance regarding application possibilities, dimensioning, selection and experiences during operation. The available literature mostly deals with answers to specific questions, like for example pressure losses in various ramifications, etc.
(ed. wgeisler - 31/5/2017)
<Blank Space>

Fig. 4: Fuller Konus-splitter                     

Fig. 5: Four-fold splitter for slaked lime in a power plant                        

1.2  Mode of Operation of Dilute Phase Splitters for Powders

For dilute phase splitters and ramifications discussed in this article, Lempp [5] had already found out in 1966 that it is necessary to segregate the particles and the gas and to calm down the solids, before mixing and distributing them again. Fig. 6 shows the basic strategy in distributing gas-particle mixtures, with the example of a patent drawing (K.-H. Mohr, 1988, patent specification DE 3626983 C2). 

In his examinations Lempp has shown that it is not a satisfying solution to use separate plates or open branches. For example, when entering the splitter with uneven distribution concentrations of the particles across the tube diameter (formation of strands) this uneven distribution can be observed behind the splitter as well. Only when using a standard T-type element, where the particles were slowed down to a complete halt after the perpendicular impact, a relatively even admixture and distribution of the solids was measured.

This is the reason why so-called riffle box splitters work deficiently, despite a high number of sub sheets and subdivision of plots, if the particle distribution in the feed is uneven. This shows that there is a systematic fault. The splitter shown in Fig. 3 will also not work with adequate accuracy, since the occurring strands will preferably charge a single outlet.

Fig. 6: Work steps of splitters [after Mohr]                                                                                         

1.3  Requirements for an “Ideal” Splitter

In order to limit the varieties of splitters and designs, only dilute phase splitters with a maximum load of up to 15 kg solids per kg of gas and a maximum grain size of one mm are discussed in the following section. The perfusion of the splitter is from bottom to top in perpendicular position. An ideal splitter has to fulfil the following requirements:

● Separation of gas and solid
● Levelling out and homogenization of the gas flow
● Strands should not penetrate the whole splitter body
● Proportioning of the homogenized gas-solid stream

Since a complete separation of solid and gas is difficult to achieve, a reduction of the gas velocity to the rate of descent of the solid particles is aimed at. In order to achieve this, the cross-section of the splitter is expanded accordingly. This expansion is continued up to a certain length, in order to dissolve strands and achieve an equalization of the mixture. Only after that, the separation of the mixture is done, according to the number of outlets.

1.4  Splitter Layout - FAQs

Is there an optimum number of outlets?
The number of outlets is based on procedural requirements. In this type of splitter, up to 16 outlets have been realized in parallel. However mostly 4 to 6 fold splitters were realized, since in this case the individual lines are less likely to be obstructed, due to the occurring reset forces. In a dual distribution, for example, the gas velocity doubles due to congestion, thus significantly increasing the static pressure as restoring force on the plug. In a 16 fold splitter the velocity in the remaining 15 outlets only increases by barely 7%. In this case the restoring forces are correspondingly lower.

How can the obstruction of individual outlets be detected?
The fastest and most reliable way to do this are pressure measurements before/after the splitter. The conveyor lines that are flown through normally have higher temperatures than the obstructed ones; therefore it is also possible to conduct temperature measurements after the splitters for detection of disturbances. Flow rate measurements in the conveyor lines behind the splitters, which are based on electronic procedures (electrostatic noise, microwaves) are also suitable. Based on the chosen comfort, these methods allow a qualitative statement on differences in flow rates.

Which accuracy can be achieved with splitters?
In case the flow-rate error refers to the theoretical mean value for each branching line after the splitter (example: total flow of solids 1000 kg/h; 4-fold splitter; mean value 250 kg/h), and an inappropriate splitter type, insufficient strand dissolution, too many outlets, changing solid material characteristics or slant inflow, deviations of up to ± 35% can be seen. However, if the prerequisites are optimal, the distribution error can achieves less than 5%, and the range below 10% can usually be achieved. In the above mentioned example this means that the conveyed amounts per line lie at a minimum of approx. 235 kg/h and maximum of 265 kg/h. Lower accuracies often are sufficient, since the main attention is on the distribution of the total flow rate as such.

Is it possible to inject additional air, and if yes, where?
Due to procedural reasons it is sometimes necessary to increase the amount of air behind the splitter, either to reduce the load (and thus the line pressure loss) or to better adapt of certain geometries of pipes. Due to the increase in turbulence, the additional air can then contribute to equalizing, however the splitter has to be designed correspondingly longer (higher).

Can different quantitative distributions of the flow rate also be achieved?
This is possible within certain limits, for example two lines with a total of approx. 40% mass flow were impinged from a 4-fold splitter, the other two were impinged with a total of 60% mass flow. The outlets at the splitter and the pipe cross sections behind the splitter were adapted accordingly. The required distribution was achieved.

Where should the splitter be positioned in the course of the line?
From the procedural point of view, a division is already possible after a few meters of conveyor line. In that case the conveyor lines (and the pressure losses) behind the splitter would be equal. The most economical solution, however, is a division as close to the injection points as possible, since in that case only one line needs to be directed to the splitter, and the remaining multiple lines can be very short. In the next chapter a few other hints are given for the equalization of the pressure loss and the division of the flow rate.

How sensitive to wear is the introduced splitter?
Wear rather happens at locations, where local high gas velocity (and respectively solid material velocity) occurs. This means, the main wear areas in a splitter are the distribution header and/or the area of additional air injection. Therefore, with conveying goods that lead to high wear, the distribution header can be removed and exchanged, if needed. It might be possible to coat the inside with a wear protection. With the removable tip it is also possible to change the distribution.

Upcoming Events