Damage to a Mixed Bed Polisher 2

[solan]

We recently damaged our strong acid cation collection grid in one of our mixed bed polishers. We are unsure what caused the event, but it bent a 4" sch. 40 SS header and several lateral collectors. The forces were such that the collectors were "wrapped" around the steel support bars that they rested on creating an "S" shape.

I should say that there are severe hydro-dynamic issues with this system. If certain polishers are placed in operation together, the flow rate through each bed tends to oscillate wildly for no apparent reason. These oscillations are minimized by placing one of the beds in manual operation.

Does anyone have any experience as to what could produce that much force? My first inclination is some sort of water hammer, but I am not convinced.

Any experience with collection grid damage would be a great help.

Thanks

 

[bimr]

It is easy to overlook the tremendous forces that are present in water processing equipment. One would guess that what you have experienced was the movement of a packed resin bed. The packed bed may have moved up and down with a piston effect. The resin bed is normally loosened up during the first step of the regeneration cycle by an upward flow of backwash water. Any channelling or air bubbles are removed together with any suspended matter which may have accumulated. The resin is then allowed to settle.

Without further details and information on your actual process, this is the best guess.

 

[cub3bead]

Solan,

You said "strong acid collection grid" but I suspect it is actually your mid bed collector and that it sees both spent acid and caustic. Is this correct?

bimr is correct about the bed pistioning upwards and damaging the collector. I suppose that you had a regen sequence upset, most like during the backwash step where the backwash water spiked as you were starting the backwash to separate the anion and cation resin beads. This lifted the whole bed as a plug and would have damaged the collector.

Even a few psi of differntial pressure spread out over the vessel cylindrical area will create huge forces on these collectors. I have seen 6 & 8" I beams trashed in this situation.

You might want to review the control logic and failure mode on the backwash inlet valve to make sure that you don't repeat this.

Good luck.

At least you had a metal lateral and collectors. Too many people buy PVC vessel internals.

The good news is after this is fixed you will get much better effluent water quality and should see longer service runs.

 

[solan]

The grid that was damaged was the lower of the two chemical collection grids. It is also just above the outlet grid (this grid is anchored to the bottom of the tank, so it wasn't damaged).

I had figured that we were pistoning the bed somehow, but seeing that both of you have focused in on the backwash step as a likely contributor, I will have a closer look at that.

I wish I could pass along all of the information that I know, but I can't keep it all straight myself. I can tell you that both events were preceded by short periods of high operating flows. The cause of the first one (so far at least) has been attributed to a very large flow experienced just as we were putting the bed into operation after a regen. It had been out of service for several days an (due to a overly large conductivity sample line) had drained down somewhat. When the unit was put into operaton, the flow rushed in until it filled the vessel. It then suddenly stopped (because the outlet valve was still closed). I figured that this caused some sort of water hamer that caused the bed to crush the laterals. We fixed the line size problem and I haven't seen that issue again. The second event also had a high flow associated with it, but I am not as sure that it is the cause.

Any more thoughts?

Thanks

PS If you want pictures let me know. They will realy twist your noodle.

 

[jmj5152]

The resin hitting against laterals would probably not do this damage. I have seen the laterals get damaged like this due to resin expansion/contraction. Resins expand and contract during service/regeneration, and can exert tremendous forces during the exhausted state. If the resins are older, you may find the crosslinking has loosened up, and the resins expand more than when they were new.
A detailed log of the regeneration chemical levels would help.

 

[solan]

The first time this happened we flushed the entire bed down with the waste chemicals (one of the laterals broke off). The second time it happened the resin was less than 6 months old.

 

[jumbowiffy]

I hate to disagree with the previous answers. However if the unit was a packed bed then piston action is unlikely, the bed is packed! I would get a sample of the resin and have it tested, if the there is any foulant that alows the resin to stick together then you might see a piston effect during service which would damage any distribution system halfway up the vessel. Also look for high DP accross the unit in the service prior to the failure.

 

[solan]

Jumbo,

We do have problems with bio-fouling. I can't say if that was part of the problem or not because we flushed the evidence through the laterals. However, if I had to guess, I would say that bio-fouling had something to do with both incidents.

What I really can't figure out is why the lower laterals were damaged and not the mid-level ones.

The only thing that I know is common to both incidents is a high inlet flow rate. What the mechanics were for changing a high flow into tons of force, I don't know.

If anyone wants pictures, feel free to ask. I will post them on a web page.

 

[cub3bead]

Jumbo,

Please go back and re-read the topic, the question relates to damaged headers in mixed bed exchangers, not a discussion of packed beds issues.

 

[jumbowiffy]

Point taken, but the comment re fouling is still relevant. Resin on the move will flow around headers and laterals if it is in its normal condition, however if it is fouled it can move as one lump, ask any resin manufacturer. This can happen to any resin bed packed or not. I have seen a mixed bed unit recently in the uk where this happened with no effect on the Cation or Anion units prior to the unit, but the MB laterals were totally wrecked in the Caustic collector system, ie the middle distributor.

 

[bimr]

solan,

Is it possible that the upper headers are more stout in construction than the lower internals?

Pictures would be great, I would like to see them.

 

[cub3bead]

I have seen beds piston when the backwash inlet valve opens quickly and causes a water hammer.

 

[sbush]

I have seen this numerous times. Usually immediately after a vessel has been drained for some kind of service. The contractor/operator, not knowing any better, simply opens the services line without first refilling the vessel with water. This sets up ideal conditions for hydraulic ram and the end results are: flat plates deflect, welds tear away, pipes bend and break. In your case, the intermediate collector was damaged by the compressive force of the water acting on the media.

Never place a pressure vessels of this type in service without first filling it completely with water. The filling process must be performed slowly until all of the air has been purged from the vessel.

It is important to have an air vent and valve at the highest point of the vessel (and piping system too).

Mixed beds introduce air for mixing in the later stages of regeneration. It is extremely important that your regeneration regime refills the vessel with water immediately after air mixing and prior to flushing. You cannot rely on a timer to completely refill the vessel: you must have positive verification of the water level by a reliable sensor device that will feedback to the operator or controller before advancing to the next cycle.

I have also seen some systems in which the water level inside the vessel will drain down during the rinse cycles because the discharge drains are located too low. If this is happening in your system, modify the drain line(s) so that it travels up to an elevation at least as high as the top of the vessel before turning down to return to the open drain. This will ensure that the tank is full of water immediately after the safety rinse is completed and before the unit is returned to service.

The oscillation of flow that you are seeing is most likely caused by air in the system. As you know water is not compressible; however, when air is introduced, you immediately have a hydropnuematic system. You need to find out if air is being introduced by a leaking pump seal or something similar and fix it; and/or install automatic air release valves in the high points of you piping system.

S. Bush

http://www.water-eg.com

 

[sbush]

Solan,

If the bending force occurred during backwash, the header and laterals would be bent upward. If the force occurred during fill, the deflection would be downward. Can you tell us which direction you observed.

Based on a 4" collection header, I would estimate that the diameter of the vessels is at or about 120". For a 10' diameter vessel the plan area of the resin bed (i.e., piston) is +/- 78.5 sq. ft. or roughly 11,304 sq. ins. Therefore, at a typical operating pressure of 60 psi, the force acting on the media bed is approximately 678,000 lbs. That's nearly 339 tons. That's almost as powerful as our 400 ton press brake machine, and I know it can easily bend 4" sch. 40 stainless pipe.

S. Bush

http://www.water-eg.com

 

[solan]

sbush,

Wow, thanks. Many of the things that you mentioned in your posts had occured to me in the past, however I had never been able to put them together before. A few items of note that occured to me while reading your advice:

1) I am positive that the first incident occured when the inlet valve was opened wide after a regen cycle. The problem was that the tank had somehow lost water while it was sitting idle (it had been sitting idle for a couple of days after it's regen). The flow into the tank went over 600 gpm before coming to an almost instantaneous stop. Sounded like water hammer / hydraulic ram to me. We fixed a conductivity sample line that was oversized and constantly running. So far, that exact event hasnt occured again. However, I haven't installed the liquid sensors or the air valves that I wanted to. I will have to make that a priority.

2) The other event, as you know, was preceeded by that odd flow problem. Air accumulation is an interesting theory, and one that is cheap to test. I will look in to that. I can definitiely see a scenario where air would produce results like that.

FYI:
The lower laterals were damaged. The mid level laterals were fine.
Both events showed definite evidence of downward force.

Thanks again!