If you are conveying in high dense phase with plugs going through the conveying line they will only break in vertically down section if the conveying velocity is less then the acceleration caused by the gravity. If it is not in plug flow then blocking will depend on the %age vol fill of the conveying line.

As far as control is concerned when the system sees high pressure it should not stop the feeder it should slow it down to achieve the set point pressure . It is a simple PID loop no rocket science in it. Then if the pressure still exceeds the set high pressure only then it should go into clear mode. ■

Dear mr Meyers,

In this forum, we had a rather extensive discussion about the definition of “dense” phase or “dilute” phase conveying.

As you state, that the system is “dense” phase, the system parameters must be known and therefore also the chance of choking. (The denser the more chance of choking)

The variable orifice controls the SLR based on the pressure, probably together with the rpm of the rotary feeder.

The variable orifice opens in case of high pressure, thereby diluting the mixture and, as a by effect, increasing the velocities, (May be not wanted in plastic pellets conveying)

Better should be to decrease the feeding in case of high pressure.

In a downward pipe section, the pellets will fall faster than the air velocity, but are also decelerated by collisions and friction..

In that section the local SLR will be lower than average, resulting in a lower chance of choking.

Plug forming or high local SLR also occurs in bends, but the airflow will prevent a total blockage and starts conveying again as soon as the remaining cross area increases the air velocity sufficiently.

Rotary lock feeder leakages can interact with the material flow in the hopper as a function of the convey pressure, causing erratic conveying. Especially when the rotary lock is worn out.

best regards

teus ■

Dave,

In my opinion vertically-down pipelines after a long horizontal section must be avoided. In horizontal lines plugs continue to grow longer and longer. These long plugs will fall to the bottom of the vertical pipe, lose their velocity, block the entire cross section of the pipe, and cause very high pressure fluctuations and system instability. For PE pellets it is much better to use a sloped line with an angle from the horizontal less than the angle of slide of the PE pellets. A slope of about 30 degrees should be sufficient for most PE pellets. This will maintain the integrity of the plugs as they travel from the horizontal section down to a lower level.

Regards,

Amrit Agarwal

Consulting Engineer

Pneumatic Conveying Consulting

Email: polypcc@aol.com

Ph and Fax; 304 346 5125 ■

I wouldn't worry about going down in a conveying line. The plug will reform once it reaches the horizontal line. If it is dilute phase, nothing will happen, you will not even notice.

Regards,

Marco Flores ■

Long vertically down conveying is often done in waste encapsulation and mining industry. I am sure some one must have experiences of such systems! I don’t see any problem with it in lean or dense phase as long as velocity and suspension density ranges are within the norm.

There is some published work on this topic also which suggests that long downward verticals can give negative pressure drops. “Let the gravity work for you”. ■

We have used capsule conveying, up and down, in our melt shops for years. We have also used dense phase conveyor, up and down, lean phase conveying, up and down. Everything works fine, except inclined lines. We tried 13º, that is the conveyor structure slope angle; we tried 45º, failed miserably; that's when we decided to go straight up, straight down and straight horizontal!

We even tried all kinds of T-bends, long radius bends, short radius bends, double T's (this was meant to fail, but the person that asked for it was a friend of the company's President. All this was cold, hot, large particles, small particles, the works!

You can find the report in the IFPS, minutes for the Pittsburgh meeting in 1998.

Regards,

Marco A. Flores ■