DI water across RO Membranes question

[PurewaterGuy]

I recently was experiencing high TDS readings on the permeate side of some RO membranes. The TDS readings were high on six RO units. The original setup was as follows.

Two carbon units prior to each RO. (that's all)

The new setup I installed was as follows.

One 5 micron prefilter, two carbon tanks, one deionizer tank, one .35 micron postfilter.

Since installing the additional equipment the TDS readings have dropped down to <1

However, I received email stating the rejection rates are dropping on the membranes.

Any ideas on whether the DI is having a negative impact on the RO membranes? Even if the PH factor is being altered it should be in the range of tolerance for the membranes.

 

[bimr]

Perhaps you have less rejection of a smaller TDS. Hoever, it still would result in better quality water.

Not sure what you are doing, but it makes little sense to use an RO process after a DI unit. The RO and DI units more or less do the same thing.

Not all water contaminants are rejected in the same manner when using an RO system. The divalent cations (hardness) are rejected with slighty higher efficiency than monovalent cations (sodium). Note that the DI effluent will containe only monovalent cations.

 

[PurewaterGuy]

The RO units were experiencing very low rejection rates prior to adding additional equipment. These RO units are being used for individual stations treating Dialysis patients. The only pretreatment called for by the company manufacturer was carbon. (No prefiters or anything else). I found this to be strange, they must have done a water analysis prior to purchase but at this point I don't think they did. So in reality, I had the two small carbon units acting as prefilters as well as Chlorine/Chloramine removal tanks. The TDS values were high and I suspect this was the cause of premature poor rejection rates. I imagine they were being fouled. I added prefilters to filter out the turbidity to a point. I added the DI tank to remove the Ions. I used a mixed-bed tank.

 

[quark]

One reason, as suggested by bimr, is that there is much less to reject after the deionizer. However, you should check the pressure difference between the feed and reject and compare this with earlier value. If the feed and reject pressure drop is more, now, than earlier then you have fouling. If it is less then your feed water quality is better.

Also check the pump discharge pressure.

As you installed a 0.35 micron filter, the feed water SDI might have been improved and membrane fouling reduced.

 

[bimr]

You probably would be better off if you installed a water softener prior to the RO instead of the DI.

The water softener will help to prevent fouling of the RO by the calcium and magnesium (scaling tendency) as well as some of the other contaminants that foul RO membranes.

 

[PurewaterGuy]

I realize a water softner is the answer to the problem but there are circumstances. All the pretreatment equipment will be tossed out the window in a few short months. I have been awarded a contract to completely pipe the entire ward and supply all the necessary equipment for pretreatment, reverse Osmosis, and post treatment as well as a emergency back-up system. The single station RO units will be removed in other words being replaced with one massive single feed system.

I used the mixed-bed DI currently because I knew it would have the same or similiar effect as a softner. However, I must admit since this is the very first time I'd ever considered placing a Mixed-bed DI tank prior to an RO I wasn't exactly sure what the outcome would be - other than better quality water of course.

I apologize, I didn't mention previously the .35 post filters were added a few months earlier. As you can imagine the membranes were being subjected to raw feed water and the carbon tanks placed prior to the individual RO's. Low flow was somewhat of an issue but not bad enough to cause the machines to alarm. BUT, I still experienced a high TDS value even with the sub micron filter, then again, there was nothing there to keep the ionic quality in check either.

 

[PurewaterGuy]

Still having a problem. The DI tank installed prior to the RO's have exhausted after just 3 days. They are only .25cft tanks. The TDS levels have skyrocketed from 0.1 to 30

I'm being told the feed water is NOT hard water.

Any thoughts? I'm having a difficult time determining what the heck could be in the water causing such havoc.

 

[bimr]

FIrst off, you need to start with the water analysis.

Next, you need to know the effluent quality that you are trying to obtain.

Then, you can select equipemtn.

Some questions:

1. Are you boosting the pressure before the RO?
2. Is the RO rejection dropping over time or isit just low?

Assuming that you are on city water, you would need good prefiltration and a water softener in front of the RO.

 

[PurewaterGuy]

Can anyone explain what Mono-chlorine is?

 

[bimr]

monochloride Definition

a chloride containing one chlorine atom per molecule.

Are you referring to monochloramine:

http://www.lenntech.com/water-disinfection/disinfectants-chloramines.htm

NH2Cl is commonly used in low concentrations as a disinfectant in municipal water systems as an alternative to chlorination. Chlorine (sometimes referred to as Free Chlorine) is being displaced by chloramine, which is much more stable and does not dissipate from the water before it reaches consumers.

 

[PurewaterGuy]

I got the Mono-Chlorine straight from the (MWRA) Massachusetts Water Resources Authority site. I was trying to find out if there was anything different being done in the disinfection process. They have been know to elevate disinfection products to the extreme, EX: Plant attendant did not add disinfrectant one day and then doubled the dosage the next day, caused all kinds of problems of course. I have never heard of Mono-Chlorine before yesterday.

I am going to secure a water sample for analysis. I spoke to the head honcho yesterday and a water analysis was never done prior to installing the Portable RO's. They worked off the assumption the units would function as they have in a different part of the city. Same water source, however we all know how drastic differences can be from one place to another. EX: I had a RO system on the top of a hill that needed acid in the pretreatment, it was the only way I could keep the membranes from crashing - kept those membranes at 98% ,,,, On the same block at the bottom of the hill I had the exact same model RO and experienced no adverse effects from the city water - RO ran and continues to run at 98% and better - same membranes in the unit now for 7 years. - The RO that needed the acid feed had 3.6cft more carbon in front of it than the one at the bottom of the hill.

 

[bimr]

"MWRA uses an mild disinfectant called mono-chloramine that has very little taste or odor."

 

[PurewaterGuy]

Well here's a real kicker, repalced the DI tanks again on the 25th and all the water quality lights are out again, and the TDS levels have gone through the roof. I've got a real problem here.

 

[PurewaterGuy]

Do you think it's a matter of not enough DI? I've got (1) .25cft of DI in line and the TDS holds well, but it's exhausting much too fast. There has to be something in the water. Water samples are being taken today.

 

[bimr]

FIrst off, you need to start with the water analysis.

Next, you need to know the effluent quality that you are trying to obtain.

Then, you can select equipment.

Some questions:

1. Are you boosting the pressure before the RO?
2. Is the RO rejection dropping over time or is it just low?

 

[PurewaterGuy]

Took some TDS feedwater readings. 80ppm. Trying to achieve less than 20ppm on the permeate side.

Taking PH readings tomorrow.

I believe the membranes are hydrolizing without the DI in front of them.

Street pressure of 60psi to the RO pump, 30psi when RO pump is running.

The rejection rate drops immediately when DI is placed before the RO, actually that's not true, the rejection rate is probably at 100% due to the purity of the water, also the rejection lights show no rejection when fed with DI.

Without DI pre-RO the TDS values skyrocket to above 30 and the membranes slowly drop in rejection, down below 70%

I replaced the 5 micron prefilter and replaced it with a 0.35 filter then removed the DI on one of the units.

The RO permeate showed a TDS of 2.1 and the membrane rejection at 95-98%

The prefilter plugged and set off the low feed pressure alarm on the unit.

 

[bimr]

If you are not getting over 90% rejection with the RO membrane, then it should be replaced.

You should either use the DI unit or the RO unit, not both at the same time. It is not practical.

You need to install a multimedia filter to remove suspended matter.

 

[vermab]

Purewaterguy,

bimr is right, the DI and RO performs same function, that is removing dissolved solids from your feed water. Placing the DI upstream of the RO will make your DI resin exhaust faster, because you are feeding a higher feed water TDS into your DI and therefore need more regeneration frequency . Your RO at this setup is just probably reducing any TOC present in the demin water coming out of the DI. When the DI resin is exhausted and if you had not regenerated them, then the RO will then assume the removal of the TDS. You are better off relocating the DI after the RO, in this set up your DI will act as your polisher producing much purer water and saving money on regeneration chemicals since resin regeneration will be less frequent due to lower TDS feed water coming out of the RO.

If you are not getting at least 90% TDS rejection in your RO , then it is time to look at replacing them. Oxidizing agents like chlorine will destroy your RO membranes, that's the reason why you had installed Carbon filters upstream of the RO (to dechlorinate the feed water). I have a feeling that there was already a chlorine breakthrough into your membrane prior to your installation of the Carbon filters.

Hope this helps.

 

[PurewaterGuy]

I do understand why the DI exhausted. I am not a liberty to install a multimedia filter at this time. As I stated before, the equipment now being used, ALL six portable RO units are going to be scrapped once I install a complete central system with the proper pretreatment. However, in the meantime I have to supply good enough water to keep the RO's from crashing. I have enough carbon before them to protect the membranes but the TDS is through the roof. So, I put DI after the carbons to remove the ions from the water to help the RO membranes. I am going to be replacing the membranes at a pretty constant pace if I can't keep the feedwater in check until such time as the new system is installed. from what I'm hearing and seeing there is absolutely no quick fix remedy that can be employed without it costing the customer an awful lot of money before the new system is in place. We do test the carbon units on a daily basis as required by Federal law in this particular application. There are two beds of carbon before Each RO. If there is breakthrough on the first bed that particular RO is not used until a carbon change is done. I have had the carbon pretreatment in place for three years. The problem I am having with the TDS values has just cropped up in the last month or so. Another thing, all the membranes have been changed on all the RO's in the last month, and on three of them they have been changed twice.

 

[vermab]

How often do you regenerate your MB DI?
A feed water analysis is handy inorder to analyze your problem. I suspect that your DI is not utilized as it should be, e.g. not regenerated at the correct frequency , when resin is exhausted, a high TDS feed is then sent into the RO. At constant RO rejection, increase in feed TDS also increases the permeate TDS.

If your objective is only to produce lower TDS product then you should locate your MB DI downstream of the RO and be sure to regenerate resin when exhausted.

You mentioned that you have six trains, you can try this set up in one train and see what happens. You might be surprised at the results!