Pumping and Piping Problem - 50% NaOH

Instead of attaching a pdf file, use the "PrtScrn" key on your keyboard to capture the image of the pdf file, and then paste it into MSPaint. Save your new graphic file as a jpg file. You'll then be able to upload it using the camera button on the control panel.

Hello climbgdh, and welcome to CR4.

I'm an instrumentation technician so I can't give much in the way of offering suggestions. However, I can tell you how you can post an image of your pdf file.

You can't post the whole pdf file in CR4 but you can post an image. So what you want to do is to copy the part you want to post onto the clipboard and paste it to an image editor like MS Paint.

Using Adobe Reader 9.0, open the pdf file and select the snapshot tool . Select the picture you want to copy. No need to press Ctrl+C (copy). It's automatic. Next, paste it to MS Paint and Save As a jpg file.

Finally, select the Insert/Edit Image tool from the CR4 editor, select your saved image and you're done.

There was a time when our production people had the same mentality as your own production people. Unwilling to provide time for proper maintenance or modification to correct problems. When things got to the point where our output really dropped to record lows, a team of experts arrived and declared what had been obvious. Now, we're given time to do things properly. It's a very welcome change.

Personally, I don't care if it's going to be decommissioned in 12 months. With the amount of downtime you're experiencing, you're probably losing more money firefighting than if you take some time to fix things properly and enable you to run trouble-free for the rest of the year.

If you can present your case in monetary terms (how much you're losing now if you continue vs how much you'll save if you shutdown to make the necessary corrections), I think management will listen to you.

I hope things work out for you.

regards,

Vulcan

It is difficult to answer without the PID and the piping isometric. You can consider posting the same to bvrao46@gmail.com Probably we can help

Top marks on a comprehensive description of the problem, and well done for registering on CR4. Welcome. Someone here will have an answer!

Very good description, as others have pointed out, but obviously a sketch will help. The only issue I'll answer now is related the the eccentric reducer. The standard is to mount this "top flat" as this reduces the possibility of forming air pockets in the suction line, and this should be complied with as far as possible. However, when the suction piping is short, installed level and connected directly to a tank, any air bubbles that form should be able to find their way back into the tank. If this is the case then the eccentric reducer may not be contributing to the problem. You mention entrained air - is there any way in which the air can be removed or is it an inherent feature of the process; and are self priming pumps used? Of course this answer, and answers to your other queries, are subject to seeing the sketches or, even better, the P&ID or isometric drawings as stated by other posters.

Very good description of the situation. In my minds eye I don't see the reducer as a problem, or I would suggest trying a few other areas.

My concern is the temperature and concentration of caustic you are handling. Is is very close to the steep solubility curve for caustic solutions. Many places switch to 40% in winter.

One question is do the solid bowl Bird's have shear pin problems and do you monitor the torque between the bowl and the scroll? This could give a clue to the amount of solids at that point.

Is there heat applied to the piping or is the solution cooling or heating to ambient conditions?

When you say salt % solids - 2 - 3% (salt crystal) Is it solid caustic (NaOH) or salt NaCl)?

I suspect that you have much more of the caustic precipitating out where you don't expect it. If this is happening than you may not have consistent performance making it very difficult to understand the performance and troubleshoot. Your sample may dissolve or precipitate solids before it gets to the lab!

I hope this helps,

Paddler

As told by Garyvan, "The standard is to mount this "top flat" as this reduces the possibility of forming air pockets in the suction line."

Occasionally, intentionally eccentric reducers may be installed upside down (flat portion on bottom). The intention is not to allow/accumulate solids (in your case salt) to precipitate at the pouch and reduce the cross sectional area of the inlet. If solid formation/carryover is inevitable and if it is more a concern than air pocket forming, 'flat portion on the top' is some times preferred.

With both suction & discharge piping are less than ideal and with upside down eccentric reducer, you have no problem, (as per your indication) "Under most operating conditions these pumps work fine and will easily pump all product from the filtrate tanks to storage."

Also you have said, "The biggest issue happens when we lose one of the centrifuges. As soon as we lose one of the centrifuges the output from the pump is substantially reduced. During this overload condition we're getting salt carryover into our NaOH filtrate and then the pump capacity is greatly reduced causing the filtrate tanks to overflow."

Under such undesirable conditions, salt/caustic solids, getting into pump is suspected to be the primary problem.

The failure continuum could be: salt/caustic solids, getting into pump > reduces the capacity of the pump > leads to over flow, so on and so forth.

How to stop overloading the centrifuge?

If two centrifuges are required in parallel for your service and one go 'out of service', you have only following two options:

1. Reduce the throughput of the system to suit to one centrifuge. Or
2. Live with the existing 'fire fighting'.

The first option is undesirable for operation and second is undesirable for maintenance group.

Is it possible to add water to the filtrate tanks during such crisis duration? Also through any nozzle at the suction line (if available). This might dissolve the solids and aid the smooth operation of the pump. Worm water (condensate, not too hot, if available) might help reducing solid formation. At 15 degree Celsius, 50% caustic starts solidifying, your operating temperature is very close by (70 deg F = 21 deg C).

Such water addition may also help loosing level during this crisis duration.

It sounds like your operators are still running both pumps even when one centrifuge is down, is that right? The new pump capacity went up from the old ones so now you empty the filtrate tank faster and your pump cycling due to air at its inlet is now faster. Why are the recycle lines not used? I have experience pumping milk of lime but not caustic, should be similar. The recycle tie point should be near the discharge tank connection, not the pump outlet, to keep the main lines flowing at higher velocity.

If you have a P&ID and some photos maybe I can help.

ron@parriottengineers.com

Thanks for the comments so far. I'll post a PID & sketch of standpipe later.

Vulcan - yes I agree operational issues are contributing to this problem. This goes back many many years (long before I became involved) and as we all know... those things are difficult to change. That being said some Operators are changing habits.

Garyvan - well aware that flat part of eccentric reduce should be on top to conform to sound piping practice. When we installed the new 2K2X2-10 recessed impeller pump recently the C/L of suction is of course a bit lower than with the 2K3X1.5-13 pump. Now from the already upside down reducer out piping actually dips down another 1.75" to mate up to the suction flange. The suction piping is short but is completely against all good piping practice. Entrained air I believe is inherent to the process. Product has a tendency to foam. There is no way we can easily remove the air. It involves significant work and no one will allow it to be done. Self priming pumps are not used. We have a 2K2X2-10 recessed impeller pump in this application. Will provide sketch & PID when I have more time later today.

Paddler - temperature & concentration of the caustic I believe is not the issue. This process has been in operation for 30+ years. The option of going to 40% in the winter is not possible. Our sales product is 50% NaOH. We do have shear pin problems from time to time on the Birds. This has been an issue for many years. We don't monitor the torque between the bowl and the scroll. If the plan was to operate this system long term the pumps would be the least of my worries. I would be recommending significant changes on how the centrifuges are operated and control. This is off the table though because of cost. No heat is applied to the piping. The NaOH comes directly from NaOH coolers and is control between roughly 21-24 deg C. By solids in mean NaCl. The level of solids increase if centrifuge is overloaded.

Yesyen - without significant changes to how the centrifuges are controlled and operated it is not possible to change the overloading. As soon as we lose one centrifuge the Operators will reduce load on the "front" end of the process. Since the centrifuges and these pumps are at the "back" end of the process the results of the load reduction will not be seen for maybe 2 hours. During that time the tanks overflow.

Orindaman - The operators almost always insist on running both pumps even though under almost all circumstances one pump is easily able to pump product away to storage. As mentioned previously the discharge piping is arranged for single pump operation (and has been for 30+ years). With them running both pumps the pumps need to be "balanced" perfectly otherwise the so called "strong" pump will recycle back thru the "weak" pump. That is another issue..... no check valves in the system.

Thanks for the comments so far. I'll post a sketch, PID & photos later hopefully.

I'm attaching a simplified PID plus 2 sketches of existing suction piping and the so called "stand pipe" arrangement that I'm proposing to use on suction side. The advantage of the "stand pipe" arrangement on is that I won't have to demolish the base and raise the pump. I've been told by our pump vendor that this stand pipe arrangement is commonly used on suction piping for stock pumps in the pulp and paper industry. Anyone here see such an arrangement in actual used. I have no pulp & paper experience so can't personally comment on it. Thanks for all the comments so far. This is a great site. I think I've figured out how to attach items to the posts... Thanks.

I think your present arrangement is better than the proposed, in respect the salt will not accumulate and will be pumped as and when it tries to settle.

On the other hand, the proposed arrangement with 'stand pipe', don't you think there is chance for the salt to settle and chock the pipe as shown in the picture?

I request the forum to elaborate on my earlier indication, "Occasionally, intentionally eccentric reducers may be installed upside down (flat portion on bottom). The intention is not to allow/accumulate solids (in your case salt) to precipitate at the pouch and reduce the cross sectional area of the inlet. If solid formation/carryover is inevitable and if it is more a concern than air pocket forming, 'flat portion on the top' is some times preferred" and comment please.

The stand pipe should have drain and vents for regular maintanance. With this, the arrangement is better.

The standpipe seems to be a workable arrangement, although I have never seen such a suction arrangement before. Good air release seems to be its goal. The suction tank elevations must force the standpipe level to satisfy the pump NPSHa. As far as settled solids, there could be a recycle line from the pump discharge to the bottom of the standpipe to maintain agitation. In fact I would think keeping both pumps operating and maintain a relatively high recycle rate everywhere would help keep the solids suspended. Are the suction and receiving tanks agitated? I would hope so.

yesyen - I'm pretty sure the standpipe arrangement would be a better solution that what we presently have. The present arrangement is contrary to many sound piping practices. We have little issue with settlement of salt in suction tank, piping, etc. That being said a drain in the standpipe would be a good idea.

b v rao - yes definitely the standpipe would have a vent back to tank plus would have a drain for clean out if required. Have you seen such an arrangement for suction piping?

orindaman - suction and receiving tanks are not agitated. because of the small size of the suction tank we have little problem with settlement. in theory the storage tank should probably be agitated but that will never happen.

All I'm trying to do is to come up with a better solution for this system that we'll only be operating for another 11 months. Whatever changes I do have to be done on an 8 hour mini shutdown. I realize that whatever short term solution I use there will be some trade offs. I'm sure we all with similar things - trying to keep a plant operating reliably, making modifications with tight time constrainsts and with limited budgets. Sometimes there solutions might not be perfect from a engineering point of view but still workable in practice. Thanks for the comments so far.

Here is a photo showing the suction piping. You can see that the eccentric reducer is flat side on bottom plus we actually dip down from the flat portion again to mate up to the suction flange of the 10" casing pump. We had no other choice to do this upon initial installation.

Your suction header is so huge. Yes, if tank looses level, air will fill and hamper pumping, even after the level builds up.

I agree your point of removing the air, preferably continuously.

I recall a similar problem, though not exactly the same.

A vertical pump in lube oil service for a centrifugal compressor train used to loose its prim often. This was due to a specific defect at the main compressor from where the lube oil picked up objectionable amount of dissolvable hydrocarbons. These high volatile hydrocarbons got liberated and collected at the stand-by pump's casing. Unless the 'air' is vented through the plugged point at the seal chamber (that is where the highest point), this pump would not develop pressure. This problem, of course, disappeared once the main compressor problem (picking up hydrocarbons) was solved. But it took two solid years for the OEM to solve the main problem. In the mean time we were to keep the fort and make this stand-by lube pump available when its service was demanded.

These lube oil pumps have 'minimum flow line' from the discharge back to tank. To restrict this stream as low as possible, a 5mm orifice was there in this line. We used this flow to eject the air from the 'vent (plug) connection'. The outlet of this ejector was quenched into tank for a sufficient depth. This arrangement worked very well and was giving interim relief until the main problem at centrifugal compressor got solved.

Adopting the same principal to this caustic pump, all you require is a small 'ejector' and simple pipe line modification as shown in the figure.

The ejector can be made using a standard ½" or ¾" pipe tee fitting as shown.

If you set up a test connection, the best efficiency of the ejector can achieved by adjusting the critical dimensions like distance between the nozzle and the jet etc. The lock nuts can be jammed after adjusting these critical dimensions.

Hope the figure is self explanatory, if further understanding is required let us discuss.

Hope this is much simpler modification and would require less time to make, when compared to 'stand pipe' modification.

Let us recall your statement, "The biggest issue happens when we lose one of the centrifuges. As soon as we lose one of the centrifuges the output from the pump is substantially reduced."

If this is the starting point of your problem, how do you think 'reversing the eccentric reducer' could solve the problem?

Do you thing the main problem is the entrapment of air at the suction pipe due to wrongly installed reducer? So the idea of installing the standpipe is to aid releasing the air, right?

"Sometimes the solutions might not be perfect from a engineering point of view but still workable in practice", well said. As long as a solution would work in a particular, existing contest, then that solution is one of the best solutions. Please consider the following point before actually going in for this modification and eliminate this possibility.

You have indicated, "Both suction & discharge piping are less than ideal". Also you have indicated at times tanks run dry. So please ensure present NPSH_A is greater than NPSH_R for the pump. Suppose if the pump is capable of pushing more liquid than what your suction line is capable of supplying, the level in the 'standpipe' will eventually loose the level and the pump would loose the prime.

Add an equalization line between the two suction sumps (use drain connections) and then add a simple float operated level control valve on the recir line to one suction sump. The other recir could be manually valved to control the flow in the equalization line. This will then automate the discharge flow to match the centriguge (combined) input. Level rises, float valve closes, level drops recir increases.

Your photo shows the worst air trap suction I have ever seen. You must eliminate this issue. Only self priming pumps could handle this suction and still it would cause problems with air if contineously received by the pumps. Keep the suction flooded, see paragraph above.