Unusual Foulant on RO Membrane System

This sounds familiar.

I had designed an Ultrafiltration unit for managing a brining operation for cheese.

It worked great for about 6 months with a steady flux rate, a run about (18 hours) at about 30 GPM and then wash.

Then after about 6 months, the Flux rate deteriorated drastically. it maxed at 28 and quickly dropped to 1.5-3 GPM within 20 minuts. Lots of investigating, misdirection from the plant management on their procedure. I'll not get into it, but basically I talked to the operators as well as the chemical suppliers that supplied their cleaning solution.

Well, their Chemical supplier (a friend of the owners family) had a 'Home Brew' cleaner that as he claimed 'as good as ECOLAB' When I question further, I found out that it was the same except for an extra polymer molecule.

The polymer molecule is added to the cleaning solution to assist in encapsulating the dirt for easier removal, usually in wash tanks and pipe lines. Problem is with membranes the molecule is a long chain, but that wasn't the problem.

The problem was, the dairy plant was using a calcium sulfate to help standardize their product. Issues occurred when the sulfide came into contact with the polymer molecule in the wash, it then turned into a gooey substance and quickly fouled the membranes with no hope of removing it.

Hence, a reduced flux rate.

Hope this helps.

^{Sometimes ones worst experience is your best, but it may not seem that way at the time.}

***** CORRECTION *****

I should proof read before the 15 minutes are up after posting... not Calcium Sulfate, It's Calcium Chloride.

Alginate gummies...

http://www.ift.org/~/media/Knowledge%20Center/Learn%20Food%20Science/Food%20Science%20Activity%20Guide/activity_alginategummie.pdf

Exactly, that's what I was thinking... a perfect storm type of deal.

15-20 years ago, I wrote a what is now known as a white paper on my experience, one of my references had similar pictures at the above link.....

pulling memories when I had my detective hat on, the cheese industry usually uses Aqueous Calcium Chloride that was purchased in liquid form, this client used a powder form. he started using it in the milk intake, but the powder never really dispersed, but clumped up. It kept knocking out the cream separator out of balance and replacing the bearings was often. This they (the client's management) never told me. one of our installers that we has sent out there to repair it about 8-10 months prior, told me that. It was another piece of a puzzle.

They kept on moving where they adding the powdered Calcium Chloride down stream, it kept rising hell with the process. The operators on the floor finally was adding it to the balance tank on the UF that was cleaning the brine operation.

This wasn't what the management told me.

they were going on their 3rd set of membranes in 30 days, on our dime... fortunately, this was uncovered before they were installed.

Yes, for me,... memories are good.... no matter how flawed.

How come it had no effect on membrane feed differential pressure, just productivity? Thin film????

Like I mentioned earlier, I'm drawing on my memory, (it was in the mid 90's). I actually have to look, I think what they did was increase the feed DP to try to keep up or maintain productivity output. But to no avail.

If it wasn't that, they also could have tore up or destroyed the membranes where the permeate just passed through. That's unlikely with an output of less the 10% of design output.

http://www.genesysro.com/genesys_spanish/uploads/docs/The_safe_use_of_cationic_flocculants_with_reverse_osmosis_membranes.pdf

Thank you, SE, for that link! It suggests (what we already knew) that cationic floc agents can be used upstream of an RO with good success when SDI (silt density index) is high (we used to be on surface water more than seven years ago), and we tried a product from one of our vendors injected into the flow of the media filters. This apparently only had a minor effect on SDI even downstream of 1 micron nominal prefilters, and did not negatively impact membrane productivity.

Later on, the water has been sourced from well fields, and has a ridiculously low SDI of about 1.5! This does not mean that the city water utility has ceased using floc agents, and they may have had carry-over events, that I will attempt to investigate.

we experienced a sharp rise in net driving pressure (all vessels), on one bank of our Reverse Osmosis system.

How is the pressure rise controlled to have the sharp rise across all the membranes?

Did you check for blockage on the membrane feed? loose gasket, buildup, Chevrolet Cavalier ????

Or on the permeate, do you have clear elbows, sight glasses or sampling cocks, if you do, is there bubbles in this permeate? Usually a sign of a failed or collapsed membrane

The pressure distribution is normal along the feed channels for these spiral wound membranes. Each vessel has six elements in series flow. Each first stage consists of four vessels, and these vessels are hydraulically in series with the second feed stage. All permeate comes off a single header pipe, at considerably lower pressure.

Normal transmembrane pressure is about 100 psig. We are seeing double that.

Normal differential pressure (feed-stage-reject) if about 10-15 psid per vessel, we are seeing precisely that, as though the membranes were brand spanking new. We have been forced to reduce throughput on this system (2-pass RO system) from 100 gpm at the 2nd pass permeate outlet, down to 75 gpm to keep from tripping on upper pressure limit of 300 psig.

wow,... Have you contacted the your membrane supplier?

What I failed to mention earlier, I had talked to ours (At the time it was AMT, Then PTI it's something else now). When I shared my information and data with her, she's the one that pointed me in the direction that it has something to do with the polymer additive in the wash.

Oddly enough, we have been considering that someone put a red rag in all of the permeate tubes, but that is just silly.

That's something I would say after I installed new membranes... "Has anyone seen my rag I had?"

Only to find out, the installers did thought they left one in there.

Seven stages of anguish (and grief) for someone that works with membrane systems.

This is usually followed by $$$!

I feel for you because your gracious enough to post the short list.

It took me 2-3 weeks to figure it out with all the misinformation the client and their chemical supplier was telling me.

if I recall, our unit was a 60 membrane unit, the first membrane replacement we install on our dime at $500.00/membrane. That's a big dime.

If the differential pressure is unchanged but the inlet pressure has increased, doesn't that mean the outlet (product) pressure has increased? Or is that too simple?

True that the last membrane in the system (at the concentrate outlet end of stage 2 vessels) is also seeing about double the normal pressure. Usually we see A1 (water transport coefficient) in the range of 9.5-11.5, now we are seeing 4's and 5's. Very disconcerting.

OK thanks I can't really picture your system without a flow diagram, but I wouldn't be able to add anything to others' posts.

There is this:

http://energy.gov/sites/prod/files/2013/10/f3/ro_optimization.pdf

for a quick overview of RO systems. Our is double redundant (A bank and B bank), 2-pass, 2-stage/pass system each. Each vessel contains six 40" length membranes of 8" diameter. The configuration is 4-2 x 2-1, 117 gpm standard rating first pass, 100 gpm second pass. Overall salt rejection is 99.99%, if not better. It still requires passage of the RO water through an EDI system to further purify the water down to >9 MΩ-cm purity or better to maintain sufficiently low trace residuals in the HRSG feed-water, and to prepare water good enough for direct GT injection.

If it's only one bank affected, could it relate to a pre-membrane filter in that line that has failed? (Allowing other contaminants into final stage.)

(Recongise that you mention this one has had more "duty" than the others for a prolonged time. I understand this also means this set has processed more volume in that timeframe.)

Could it be the membranes in that set have been drained and left dry for a period? (Though this would probably provide different symptoms, especially in quality of produced water.)

Is one membrane in the parallel set totally blocked and the pressure represents the flow going through fewer cartridges? (Stuck valve, broken valve stem, actuator uncoupled from valve shaft on rotork.)

Just tossing in some other ideas.

A lot of things have been checked:

1. Prefilters upstream of this bank are showing only 6 psid (in nominal range).
2. Flow totals of permeate produced with concentrate compare well with flow through media filter.
3. Pressure gauge calibrations were done only last week, and all of the transmitters were within 1-2% of the calibrator.
4. The membranes in this "B" bank (of the 100% redundant system) are newer by far than the ones in the "A" bank, and "A" bank is highly productive, although salt rejection is starting to decrease slightly with age.
5. Low pH, high pH, and surfactant cleanings all produced negative results. No pungent organic odor or sticky material in foam during cleaning suggests that there is effectively no bio-fouling on "B" bank.
6. Salt rejection on "B" bank is excellent, in fact almost abnormally good.

Del's got wonderful formula for cats.

Troll alert!

Trolley Alert!