Desalination/Reverse Osmosis - High pH Permeate

In theory, the pH of Permeate produced from Desalination of Water should be lower than the Feed Water

Meaning acidic

Since its a higher (more than 7) its a base or alkali.

http://en.wikipedia.org/wiki/PH

(sorry about laying out the basics)

Whats the (MWCO) molecular weight cut off of your membranes.

Should do this systematically but a quick response could there be a possible contamination point to the process, or a failed membrane?

Whats the PH of your retenate?

How many gallons or liters a minute is the premeate, and what is the flow over the membranes?

Is this a continuous RO process or a batch?

phoenix911

Hi Phoenix911,

Yes you are correct, the projections indicate a slightly acidic (low pH) permeate, however we are observing a slightly basic (high pH) permeate.

The MWCO isn't specified by the manufacturer, but the membrane has a 99.75% Salt Rejection. Membrane is Dow/Filmtec SW30HRLE-400
(link: https://www.dow.com/liquidseps/prod/sw30hrle-400.htm )

The poly-amide menbrane strongly rejects ions, smaller non-ionic organic compounds pass freely (with no rejection) through the membrane, small ionic compounds such as Ammonium have about 50% rejection. Larger organic compounds are more strongly rejected (ionic more-so than non-ionic).

We've tested the conductivity of the water from each vessel, all were the same - we don't believe there is a failed membrane or seal.

pH of the retentate is virtually the same as the feed (approx 7.86 - 7.95pH).

Average flux through the membrane is 14.0 L/m2/hr, the minimum cross flow velocity across the membrane surface is 4.5 m3/hr.

Continuous RO process.

Thanks for your help, very much appreciated.

Anthony

Are you using Carbon before the RO, and if yes when was it last replaced

I've encountered this problem before. In my case it was due to a higher than expected dissolved calcium level. It can be easily resolved by dosing in some hydrochloric acid.

Hi DVader1000,

Thank you for replying, it is comforting to know someone else has atleast seen this happen.

But in this case we are experiancing even less Calcium than expected (Water analysis shows 366mg/L in the RO Feed, we were expecting approx. 420 mg/L). Do you know what level of Calcium you were experiencing ??

The Water Analysis of the permeate showed the Calcium concentration was undetectable (< 0.1 mg/L). This is understandable, as I would expect effectively complete rejection of Calcium by the RO membrane.

Do you know the chemistry that caused the high permeate pH in your case ? Perhaps the theory is transferrable to my situation.

Any assitance is greatly appreciated.

Kind Regards,

Anthony

I suspect it's calcium contamination in the feed water tank in my case. The amount of calcium by analysis was actually fairly low, only 12ppm or so, but my 2 clues are that the pH value kept going up, from a low of 9.0 in the raw water tank, to a high of 10 by the time the water enters the RO pump. By the way, the raw water used in my case was actually RO permeate.

I didn't manage to get the reject analyzed as the client had stopped operating the unit for some time due to scaling, but I suspect the pH could be as high as 11.0, since my own analysis of the scaling indicated that it was calcium carbonate.

• How old are the membranes?
• When was the last time they were chemically cleaned?
• Are all the inter-membrane connectors all both mechanically complete and watertight?

Assuming that the RO filters are in good condition and there is no post contamination, the answer should lie in the composition of the estuarial water. You don't state that you have measured the CO₂ concentration of the incoming water; if not, this can be depleted relative to sea water by photosynthesis - and this would not normally be a problem. However, river water can also contain* heavy acids (H₂SO₄ upwards,which you will be pleased to be rid of) and light alkalies (e.g. various Li compounds, which may be of concern if present to excess). Either of these would cause the effects you observe, and (being conservative by nature) I would be inclined to send a sample for more comprehensive analysis

*Depending on type, that could be be human/industrail contamination or geological

How long has the condition existed? Was this a slow drift in changing ph values? Or was it an abrupt change?

Do you track the ph values of the feed water? Has it changed over the same period of time?

This is a longshot but...

What is the conductivity of the water that is giving unexpected pH results? (compared to what the meter needs to read)

I have seen very pure (low contuctivity) water play havoc with the results of pH probes. Had to add a salt to bring conductivity up to where the probes can read.

This was not with RO but with a multi step Ion exchange resin train.

Paddler

having done work in the pharmaceutical industry. . . .purified water is corrosive

In wastewater RO applications, pH was adjusted from above pH 7 to about 5.8 (by injecting sulfuric Acid or HCl) to reduce scaling. RO Effluent is processed through an air-stripper/degassifier to remove H2S and carbonic acid ... the result was an increase in final pH.

Also Note: Saline pH probe/meter calibration may be different than non-saline water calibration. If you are relying on the same probe/meter and calibration procedures for influent and effluent, it may skew the results. Verify your readings with grab-samples and a separate hand held meter.

I hope this helps.

Hi All,

Firstly I'd like to thank everyone for thier input, it is all very much appreciated.

Replying to a few questions;

- Plant is not using Carbon Pre-treatment. The only pre-treatment is a 100 micron strainer and Micro-filtration plant.

- The menbranes are brand new, I'm still commissioning the plant. Just looked at the control screen they've done exactly 85 hrs, 59 minutes run time.

- Membranes have not yet been chemically cleaned. The RO Plant was mothballed for approx 3 weeks (client had serious power failure), they were stored in dilute SMBS solution.

- Regarding the intermenbrane connectors, to test this we measured the conductivity on each vessel individually, the results are below (I hope the table shows up OK).

With the error column, I took the lowest 5 values, made that the benchmark, then calc. the error of each vessel from this number. The highlighted are greater than 10% different from the ideal.

I've discussed with people about this, and they suggest that this isn't too bad, there are a few to be concerned about, but as the plant is producing water within conductivity specification. They also suggest that in thier experiance leaking seals, and broken membranes would not cause increase in permeate pH - anyone have thoughts on this ??

- light alkalia's (Lithium), and Heavy Acids both are good suggestions I will get further analysis, and research these.

- CO2 levels were measured. I will post complete analysis is separate post.

- This is a recent observation, but the plant is still in it's infancy (in commissioning). The readings appear to be consistantly greater than expected(although they vary significanty. We are logging data, and over the next few weeks we'll see more).

Thanks all for your suggestions, please keep the dialogue coming. This has many people perplexed. When we find the solution alot of poeples minds will be put to rest.

Kind Regards,

Anthony

Leaking seals and damaged membranes can cause a higher than projected permeate pH, because some of the cation-rich reject will leak through the gaps and contaminate the permeate.

Hi DVader1000,

You're right it certainly can. However it would also increase the TDS significantly.

I did some quick projections (blending feed with the permeate) and to increase the pH to the approximate levels I'm seeing approx 3% of the feed needs to blend with the permeate. This causes the TDS in the permeate to increase from approx 300 to 3000 mg/L.

However the TDS the plant is making is even better than projected (at less than 250 mg/L).

So even though I'm getting some slightly higher conductivities from a few of the vessels, I don't think this is the root cause.

Currently the plant is running well (except for the high pH permeate). I havent updated the forum for a while, as there was really no news to speak of.

Problem still exists with now explanation to date.

I do appreciate your thoughts on this. Please message me if you have any new idea's on this.

Kind Regards,

Anthony

Water Analysis and RO projections ...

Hello ajwinemaker:

I think you may be dissolving the tank and pipe if they are concrete? It may be just one part of a tank or the flow-way leading to it has been made from slightly to wrong ingredients.

How long has it been like this? And it it possible to fit extra post desalination reverse osmosis? I would suggest a filter element of around 5 microns or smaller, it really depends on how 'clean' of particulates the final product is at the moment.

Are the tanks enclosed and anywhere near soil? Like as in a grass bank?

Take care..............

Good answer. RO permeate pipes should be made of thermoplastic, not concrete or SS304. Even SS316 isn't really recommended.

Hi,

we are facing the same problem. Just want to ask if you have solved this problem or not.Thanks

Hi, yes I believe I understand this well now. I've also seen it on several other plants.

It all depends on your water source - can you define your water source? post water analysis if you have it?

Basically, it is false alkalinity. To determine the alkalinity shown in the water analysis, the lab will titrate using sulfuric acid to a colour change using phenolphthalein indicator.

Essentially this titration is measuring the buffering capacity of the water, and then assuming the extent of buffering capacity is entirely due to carbonates/bicarbonates.

There is some research showing that other compounds will interfere with the colour change titration giving the impression there is an incorrect amount of alkalinity in the water.

Therefore if there is no carbonate / bi-carbontates in the water, there is no CO2 passing through the membrane, then there is no carbonic acid in the permeate (hence the pH is not reduced).

I have a paper on my HDD at work I can post, next week. Basically the water sources where this is observed is rivers and water sources contaminated with heavy metals. I've seen this in three mine sites, treating mine / process water, tailings, and a SWRO plant that is fed by a river with heavy industry upstream.

Hope this helps. More than willing to assist further, please post relevant info about your water source and the RO plant.

Cheers,

Anthony

Many thanks.

I am currently measuring the alkalinity of the RO feedwater.

For the reason of heavy metal, I do not think it is the potential origin as we did not see a major change of this component in the water.

Very interesting to note is that we have a drift down of ORP together with this thing happening.

The peremeate conductivity is also rising tremendously.

Hi,

Agree - Very interesting.

Regarding the alkalinity measurment - check with the lab and if it is measured by way of titration, then be wary of this result.

Regarding the ORP - it is funny you should mention this as I also noticed great fluctuation in the ORP. As I was commissioning this plant, I had great troubles with the ORP readings. I re-commissioned the ORP many times as the instrument changed so often (frequently over the period of an hour would be way out). We really didn't ever find out why the ORP fluctated so much, after 2.5 years we autopsy'd a membrane and no oxidation was found, so I guess the ORP was doing it's job.

Regarding the Permeate conductivity, suggest you treat this as seperate issue. As this indicates problem with the membrane. Whereas the high pH permeate does not. I can assist you troubleshoot this one if you'd like. I'd start with checking the accuracy of the permeate and the feed conductivity analysers, by comparing to accurate handheld. Often the permeate condy has drifted.

If instruments are accurate, narrow down the source(s) of the high condy water. Is it from one, several or all the vessels? If only one, do a profile test (to detemine what membrane or interconnector is the source)?

Hope this helps.

Anthony