RO Prefilters 1

[ProjectEng]

Any recommendations on what to install for filtering the water going to an RO unit?

Inlet water to the filter would be potable water that varies in quality since the source is a lake.

Throughput needs to be 4,000 - 6,000gpm.

I think sand filters are out of the question because of size constraints.

This plant has had many problems with their RO because they would run straight city water to it.

 

[tr6]

Pretreatment filtration upstream of the RO membranes should consist of: multi-media filtration (antracite, sand, and garnet) to reduce suspended solids to less than 1.0 NTU followed by .5 micron cartridge filters.

The .5 micron filtration is required due to the pore sizes of the membranes themselves. And the multi-media filtration is required to reduce the loading on the cartridge filters as these types of elements cannot be cleaned; rather when around a 5-8 psid is reached across the filter it is taken off-line and the elements manually changed. With proper upstream filtration, this procedure may only have to be performed every few months or so.

You said that they are currently running city water directly into the RO. I trust there are chemical feeds to reduce residual chlorine to 0 and the addition of anti-scalant chemicals to prevent scaling in the membranes?

 

[johbil]

I agree with the previous post. Treatment should include antiscalant dosing and active chlorine removal (NaHSO3). We have been running with 5 and even 10 micron prefilters for quite some time. We basically change the filters every 4 months of use. Pressure drops over the elements are about 1.5 bar at that time.

 

[drwater]

If they ran straight City water through the Plant w/o removing the chlorine, then the membranes are swiss cheese (unless the membranes were CA rather than polyamid). Basically you are looking at two treatments; surface water and RO. With the flowrates you mentioned (6MGD to 9MGD) you could probably use a microfilter or ultrafilter membrane system (rather than sand or multi-media)and lower the overall O&M costs in the long run. With cleaner feed water to the RO system, you could afford to increase the flux rate through the membranes (to upwards of 20 GFD) and get more throughput. The key to any RO is pre-treatment. Keep the Silt Density Index (SDI-15 min.) to less than 1.0, use scale inhibitor/acid if needed and keep out uglies like TSS, Fe/Mn and Silica.

 

[quark]

As drwater mentioned you have to take care of SDI and suspended silica also. Beware of free chlorine. If at all you are treating with chlorine using SMBS dosing is better rather than activated charcoal filter.

Regards,

Repetition is the foundation of technology

 

[ProjectEng]

Actually, the RO units at my plant have been offline for some time. The fix is expensive and we want to make sure we get it right. Currently, we're running city water to the IX's. If we can get the RO back online, it'd be a huge savings.

Yes the anti-scalant was being added and sodium bisulfite treatment for chlorine was being done when the RO was online previously. They had UF filters before RO's. But the UF's plugged up (due to solids/silt in the city water) so they bypassed them (dumb) and went straight to the RO.

My concern with multi-media filters is that they would have to be too large for our flow-rate and we're somewhat limited on space. Does anyone have experience with the cartridge-type filters that can be backwashed?

 

[aspearin1]

Wow!!! That's an incredible flow. Just today we were looking at UF for a process, and decided it would be a monsterous amount of membrane to even achieve 20gpm.

I had come across a filter type the other day we were condsidering for our own use. It's an autocleaning filter, generally used on machines with recycled fuel and oil. It can continually remove sediment above a certain size and cleans itself. Check out

http://www.cuno.com

Contact the SASS group, if you are interested.

 

[ProjectEng]

aspearin1,

You're probably using very small membranes. Our RO membranes are 8" x 40" long and rated capacity of 1 membrane is over 6gpm of permeate.

 

[aspearin1]

ProjectEng,
What is the MWCO or NMWL of your membranes?

 

[ProjectEng]

aspearin1,

I don't recognize those acronyms. The membranes are Dow Filmtec BW-365, so they've got 365ft2 total area each. With our high-pressure multi-stage pumps, our RO inlet pressure is 360 - 400psi.

 

[aspearin1]

ProjectEng,
The acronyms are Molecular Weight Cut-off and Nominal Molecular Weight limit. It seems the difference between the systems we were looking at was to operate at a much lower pressure. More information on the autocleaning filters: the cleanest rating is nominal 40 microns, meaning it will catch most particles at and above 40 microns. I'm not sure how much help that is, but I thought I'd throw it out there.

 

[ConsultUS]

Halo Project Eng.
Congratulations , you made a great Panel.
Everybody share a little, to make a big total and in my opinion every body is right.
I designed a system after looking at the problems of other system and it has been running great, with waste paper mill water.
The Key issues for sucsesfull design were.
1.-Identify composition and sise of contaminants to be removed. Sorry there is not free lunch in technology, you have to analize.
2.- Do all the Chemical conditioning recomended
3.-Then do Do Prefiltration in stages, acording to the amount and sise of contaminants.
4.-Prefiltration is not one or the other, you need them all and probably an Ultrafiltration membrane before the RO membrane is required, I sugest you get into the clasification of membranes by the dimentions of the particle you try to separate
5.-If my gut feelins are right you will need first Coarse filtration,Multimedia with activated carbon will help If you have a lot of disolved organic material, as you will, if the water comes from a lake, then Ultrafiltration , and then the water will be conditioned for the High pressure membranes you have, they will will work beter
6.- aspearing 1 is in the BINGO trayectory I sugest very strongly you pursue this lead very intensively
7.- It looks like the membrane design was not the best some body did not made all the previous Homework and now you will have to do it.
Good Look.

 

[drwater]

OK ProjEng, let's start from the beginning... you say this is city water, I assumed it was already treated to potable standards, but you just wants to reduce the TDS down further. What exactly is the source water quality like in terms of TDS, suspended solids, etc. The ultrafilter you mentioned "plugging up" must be ancient. The newest ones by USFilter Memcor (microfilter) and Zenon (ultrafilters) also Pall Corp., Hydranautics, etc., are capable of removing very high turbidities (>100 NTU) w/o pre-treatment. Also, for the previous responders benefit, there are MF/UF treatment plants of >25 MGD in size. You are using IX now, which tells me you are trying to either reduce total hardness or you have ion specific resins for nitrates, etc. The Filmtec membranes you mentioned are also older technology, and newer generation brackish water (BW) membranes should be operating at much lower pressures (100 psi - 200 psi). Please provide a little more info as to feed water quality and what you want the treated water quality to be. Thanks!

 

[ProjectEng]

Yes, it is basically city water except that nitrates are not taken care of as diligently. So it's almost potable.

I'm still working on getting a 12-month profile of the water here. The numbers I'm seeing in our inlet water reports are a bit puzzling because the water quality seems fairly good. Turbidities tend to stay below 2.0, alkalinity below 1.5, TDS below 400, and SDI below 7.0.

But the RO has never performed well. There may be some design issues. One thing that strikes me is that the RO is vertical. Any comments as to how detrimental this is?

I'll check up on our ultrafilters. Our water treatment service company has already suggested low-pressure membranes and we will eventually go that route. But I want to make sure our pre-filtration is set up right.

We ultimately need demin water for 1000psi+ boilers.

 

[drwater]

The orientation of the pressure vessels does not matter. Typically, in larger systems, the orientation is horizontal. This is really for removal/replacement of the membranes, and not a hydraulic issue. Your water does not seem bad at all, with the turbidity and SDI cleaned up a bit the RO treatment should be a piece of cake. Maybe there is something else in the water, such as high levels of iron or manganese that is oxidizing and causing plugging. One thing you can do is send an old membrane (one from the front and one from the end of the array)to the manufacture and they can perform an autopsy on it (cut it open and look at the membrane under high magnification) to see what is plugging/fouling it. With that low of a TDS, even going to DM/DI water, should be easy for a two-pass low pressure RO. I am interested in seeing other water quality results as you get them.

 

[NicSacco]

Please look at the AFR on the

http://www.rpaprocess.com

webpage. Their phone number is 269-323-1313. This filter is touted as a replacement for sand filters but measures 4ft x 4ft.

nic

 

[bimr]

My understanding is that the recommended SDI before RO treatment should be in the range of 3 to 4. Since you do not have that SDI, pretreatment is in order. The least expensive pretreatment approach is to use a multi-media pressure filter to filter the raw water.

Note that it is common practice to filter potable water prior to RO. By the way, you can stack the filters on top of each other to save space.

IMHO, the older UF system should have had a multi-media filter pretreatment system. The older UF was probably installed in lieu of the multi-media filter pretreatment. Probably a bad decision.

If you install a multi-media filter and use a little cationic polymer, you will probably get a SDI of 1.

You should also consider replacing your DI unit with the newer CDI units as the same time.

The backwashable cartridge-type filters that I am familiar with have been used in condensate polishing systems in nuclear plants where the goal was to reduce radwaste. This filters are similar to the standard cartridge filters, but are backwashed to avoid having to change a hot filter. These are only used on water with extremely low solids. These filters are probably too expensive for your application.