Mechanical design of PTA/PET/PE conveying lines.

Posted in: , on 16. May. 2003 - 10:27

We have experienced quite a few failures in the primary as well as the secondary support systems of the various conveying lines for PTA/PET powder and PE pellets.

The mode of conveying is both lean and dense phase.

Some of the problems are:

1. High vibration during conveying

2. Very large displacements at changes in direction

3. Damage to pipe supports and to the structure due to high dynamic loads.

We presume that this could be due to slug flow or operating at conditions higher than the design parameters.

However, it is always the contention of the operating personnel that they are operating to the design parameters.

As we have no method of estimating the slug dimensions and therefore to calculate the dynamic forces, we end up over -designing supports, both primary as well as secondary.

Could someone please enlighten me as to how we can estimate the dynamic forces, when and how slugs can develop and if they do what are by methods by which we can break-up the slug.

In our own rather primitive way we estimate the size of the slug as ID of pipe and length as twice the ID. Knowing the bulk density and the velocity of flow we calculate the change in momentum.

At bends especially at risers we are recommending installation of small nozzles circumferentially to create a blast effect using Nitrogen.

Are we on the right track?

Can someone please advise.


Dennis Hauch - Freeport, TX, USA
(not verified)

Dense Phase Support Systems

Posted on 16. May. 2003 - 05:39


The discontinuous nature of dense-phase conveying is synonymous with the pipe forces / pipe movement that you are experiencing. These can be accommodated, however, by a properly designed pipe support system.

It is imperative that the support system allows for a controlled movement of the pipe. Pipe that is rigidly supported will certainly be broken as will the supports, primary and secondary.

Coperion Waeschle has done a great deal of research with regard to the dynamic loads that are involved. This research has resulted in a design program that now has a proven record of success. Please contact Hans Schmidt at Coperion.

Research / experience has established the relationship between slug length and the length of the horizontal run preceding the bend, i.e. the longer the run the longer the slug. Imagine the momentum of a slug 20 feet long, which in not impossible.

There are numerous designs for devices to split slugs, some are very effective, and some are not. The Coperion designs have proven to be very effective. I would not recommend the nozzle approach that you are considering. It offers little hope of splitting or even softening the slug. On the other hand, it could very easily add to the velocity / momentum of the slug.

These are my thoughts.

Dennis Hauch

Re: Mechanical Design Of Pta/Pet/Pe Conveying Lines.

Posted on 17. May. 2003 - 07:20

Mr Dennis Hautch,

Many Thanks.

I wonder if we could alter/change some process parameters

like, air to powder ratio, velocity, etc.

I agree we should allow pipe movement, both thermal as well as due to dynamic loads. To absorb thermal movements we do provide expansion bellows with arch filling (Posiflex). I am tempted to consider providing shock absorbers at all changes in direction especially when rising up to a silo/hopper to absorb the dynamic loading in the slug flow regime. Any comments?

Could you please let me know how to estimate slug lengths?

Thanks once again.


Dennis Hauch - Freeport, TX, USA
(not verified)

Dense Phase Optimization

Posted on 17. May. 2003 - 05:52


Yes, the dense-phase conveying process can be optimized.

A key parameter here is the gas velocity at the pickup point. The gradual minimization of this parameter will cause the system pressure to rise. This can continue until the pressure rises to, say, 80% of the supply pressure, or until conveying becomes unstable, whichever is first.

The above describes a very basic control philosophy. A highly refined process controller called Intacon is now offered by Coperion Waeschle, it is one of their featured attractions at their booth at Achema ’03. The advantages of automatic control are many, e.g. stable conveying, minimized pipe forces, maximized solids loading, maximized throughput, accommodation of varying product types and conveying distances.

Some allowance for movement of the conveying pipe is needed at every change in direction. This becomes increasingly critical as the terminal end of the system is approached, this, because the velocities are higher and the forces are greater.

As to slug lengths, you can get a “feel” for the length by hugging the conveying pipe upstream of a bend. Or you can actually measure the slug length with light beams, but this falls in the area of research which is expensive to do in operating plants. Again, Coperion Waeschle has done extensive research in this area which led directly to the development of the Intacon air control system. Check it out.


Dennis Hauch

(not verified)

Re: Mechanical Design Of Pta/Pet/Pe Conveying Lines.

Posted on 18. May. 2003 - 08:42


Installation of additional nozzles without understanding the nature of flow in the lines is not a good idea.

I suggest that you first try to optimize the conveying velocity in the line. The conveying velocity at each of the locations in the line should be maintained such that you do not land up in the unstable zone. This can be achieved by adjusting the gas inlet flow & may be stepping the lines. The test data from your conveying trials (if done) should be handy here.

If your conveyeying rate varies widely, you can still have problem in optimization. The gas flow rate must be adjusted as a function of conveying rate to keep the system in stable conveying region. A number of smart dense phase air control systems (mainly by Coperion and Neu Transfair) are commercially available.

Additionally, you should look at the possiblity of supporting the lines properly. Coperion, Zeppelin and Neu Transfair have done substantial work in figuring out the dynamic loads. Do not take the "rough calculation" method to guess these loads.

Plug breaking or conditioning is also a possiblity. There are plug conditioning concepts available from these vendors.

Just curious, who designed your system ? Was any tests done ? At what scale were the tests done ? What is the current scale of your system ?



Re: Mechanical Design Of Pta/Pet/Pe Conveying Lines.

Posted on 19. May. 2003 - 07:29

Jackarcher/Dennis Hauch,

Gentlemen many thanks for your valuable suggestions/comments.

Proffessional ethics prevent me from naming the consultant/vendor. As suggested we are in touch with various consultants including Coperion for the existing as well as some ongoing projects.