My AWT test results....what changes should i make?

I got a different response from SpectraPure. The response I got indicated that silica would be detected by the COM 100, at least.
 
Humm

RO plays a major role in silica removal

The study of the chemistry of silica (silicon dioxide, SiO2) and its relationship to reverse osmosis (RO) is of interest on two fronts: One is the challenge silica presents as one of the most difficult-to-handle inorganics for the RO process and the other concerns efforts to better understand the basics of RO.

Both have to do with silica's chemistry, which is unique - especially when dissolved in water. Much of this may be related to silica's relatively moderate polarity; some say it has to do with its similarity to the water molecule itself. What is known is that silicon is one of the most common elements. It has a rather low solubility and tends to interact with water molecules when dissolved. Silica is generally found in water supplies in three different forms, as reactive, colloidal and suspended particles (e.g., sand). The reactive portion of the total dissolved silica readily reacts in the standard molybdate colorimetric test, and the colloidal does not. The reactive form is silicon dioxide dissolved in water, creating the compound monosilicic acid (H4SiO4). In this form, silica is generally un-ionized at most natural pH levels. At a pH of 8.5, only 10 percent of the monosilicic acid is ionized; as the pH reaches 9 to 10, it still is only 50 percent ionized. The colloidal species is generally thought to be either silicon that has polymerized with multiple units of silicon dioxide, or silicon that has formed loose bonds with organic compounds or with complex inorganic compounds, usually aluminum and calcium oxide structures. Monosilicic acid attracts four additional water molecules beyond the two that make up part of its molecular structure in the hydrated state. The overall hydrated structure contains a total of six water molecules that probably play a significant role in its behavior in the RO process. Silicate also plays a role in alkalinity measurements since it is titratable with acid. For example, at pH 9.65, for every 100 milliequivalents (meq) of silicate, 58 meq of alkalinity are contributed to the total alkalinity. For high pH water supplies that contain appreciable amounts of silica, water treatment professionals need to include this equation in the alkalinity calculations in order to create a properly balanced water analysis. Troublemaker: colloidal silica Colloidal silica creates problems for water treatment because of its stability as an un-ionized compound, which makes it difficult to remove using ion-exchange processes. It can even cause some resin fouling where colloidal silica levels. are exceedingly high. Silica is also found at the lower end of selectivity for anion resins, creating a scenario in which silica is one of the first contaminants to breakthrough. As a result, silica can be effectively removed only if the ion-exchange resins are completely and properly regenerated. With the introduction of RO as a pretreatment process, colloidal silica can be removed effectively using a typical RO membrane. The RO also helps the ion-exchange process by alleviating the overall silica loading on the resin. The downside of silica pretreatment by RO is the effect that the relative insolubility of silica can have on RO membranes as the feedwater becomes more concentrated. For example, at a pH of 7, the solubility of silica is only 120 milligrams per liter (mg/L) at 77 degrees Fahrenheit. This presents a problem for an RO operating on feedwaters having a silica content of more than 30 mg/L when the recovery rate needs to be in the neighborhood of 75 percent. At the resulting four times greater concentration, the RO concentrate would contain roughly 120 mg/L, which at neutral pH might begin to show signs of precipitation. At pH levels that are more alkaline, however, this solubility becomes much greater. It is conceivable to raise the pH in high-silica feedwaters. However, operating most RO feed streams with a higher pH invites other problems, most notably calcium carbonate precipitation. As a result, these circumstances usually call for presoftening of the feedwater supply using ion-exchange softening. Softening offers two advantages: The pH can be elevated, and (it is believed) that the potential for silica precipitation in the concentrate stream is minimized because of the absence of calcium, magnesium and other divalent ions that could serve as precipitation nuclei. lt might also be worth noting that it is generally believed that at neutral or acidic pH levels, it is not a great risk to approach silica concentrations - on softened feedwater - as high as 150 mg/L in the RO concentrate. This is mostly due to the perception that silica precipitation occurs at a relatively slow rate and that the concentrate will be discharged before silica scale is actually formed. Also, some silica inhibitors can be added as a means of pushing past the solubility limits and avoiding precipitation. Silica and the RO process Studies have shown that the colloidal species are removed to >99 percent, even with cellulose acetate (CA) RO nanofiltration (NF) and ultrafiltration (UF) membranes. Reactive silica is rejected well by thin-film polyamide (PA) membranes, yet is not reject as well by CA membranes. Taking into consideration that reactive silica is essentially un-ionized at neutral pH levels, some of the fundamental rejection theories of the RO process may need to be reconsidered. That is, why does an un-ionized compound such as silica reject to such a high degree if the rejection of salt is typically related to valence, as is so often accepted as the mechanism for ion rejection? It is known that other completely un-ionized compounds at neutral pH levels, such as hydrogen cyanide, pass readily through an RO membrane. So, why does silica reject so well with PA membranes, even at a neutral pH? At a time when CA membranes were the only option and silica rejection was quite poor, all seemed to make sense. Silica always showed greater passage rates than inorganic species that carried a charge, even the relatively high-passage monovalent chloride ion. However, as PA membranes became available, it was found that the rejection of silica was very high, with little variation up and down the pH scale. Could it be that the rejection from RO membranes has little to do directly with the valence, but more to do with the physical size of the ion? This might begin to explain why the rather large silicon dioxide molecule rejects better than a chloride ion, even though the latter carries a much more significant charge. If the fact that most compounds may have an attraction for nearby water molecules (to differing degrees, depending on the nature of the compound) were taken into account, the notion that rejection of ionic species may be more size-dependent ("size" being the hydrated diameter of the compound or element) might become accepted. Lee Comb is vice president for business development for Osmonics Inc., Minnetonka, MN. References 1. ller, R.K., The Chemistry of Silica, John Wiley and Sons, New York, NY, 1979. 2. US Geological Survey, "Study and Interpretation of the Chemical Characteristics of Natural Water," Supply Paper 1473, US Government Printing Office, Washington, DC, 1970. 3. Jenkins, D., Snoeyink, V.L., Water Chemistry, John Wiley and Sons, New York, NY, 1980. 4. Lerman, S.I., Scheerer, C.C., "The Chemical Behavior of Silica," ULTRAPURE WATER, December 1988.


and

Titre du document / Document title
Performance of RO membranes in silica bearing waters
Auteur(s) / Author(s)
SHEIKHOLESLAMI R. (1) ; ZHOU S. (1) ;
Affiliation(s) du ou des auteurs / Author(s) Affiliation(s)
(1) School of Chemical Engineering and Industrial Chemistry, The University of New South Wales, Sydney, 2052, AUSTRALIE
Résumé / Abstract
Performance of two types of reverse osmosis (RO) membranes in silica bearing waters were investigated. The experiments were carried out using the Osmonics SEPA CF membrane flat sheet unit. The membranes investigated were polyamide (MS19) and thin film (DS11); their rejection, compaction, concentration polarization, fouling propensity and ease of cleaning were examined. In addition, the effect of silica concentration and polyvalent ions were investigated. The average silica rejections on DS11 and MS19 membrane were found to be 96.2% and 93.8%, respectively. The thin film (DS11) membrane was found to be easier to foul by silica than polyamide (MS19) membrane, but more easily cleaned. Both membranes underwent compaction; however MS19 was more prone to compaction (15-30%) than DS11 (0-15%). The concentration polarization ranged between 2 and 3 for the runs. Increases in Ca, Mg and total hardness for runs with given silica feed concentration increased fouling penalties. Analysis of the deposit by EDS indicated that Ca and Mg did not partake in chemical structure of the deposit; it appears that they catalyzed polymerization and were minutely adsorbed by the silica deposit. The results indicated that keeping low feed silica concentration might not, under all conditions, be effective in minimizing fouling. For Ca to Mg ratio of 1.5, the optimum initial silica feed concentration was found to be 100 ppm, not the lowest concentration of 50 ppm. The emphasis should be put to understand the mechanism of silica fouling (i.e. crystallization or particulate fouling) in order to mitigate this problem. Further experiments are needed to investigate the effects of other parameters on silica fouling in dynamic systems and to understand the mechanism of silica fouling.
Revue / Journal Title
Desalination (Desalination) ISSN 0011-9164 CODEN DSLNAH


IMHO special silica DI's after a RO is just a bunch of hype until some can prove other wise. :) I want to see a RO/DI assay showing silica concentrations. ;)
 
Boomer, after getting my head around the passage you posted (somewhat.. :) ), it seems like "CA" membranes do a great job of rejecting colloidal silicates but not so good with the reactive variety. I guess it's implied that the other membrane varieties work well with reactive silicate. I use a TFC membrane in my RO, so I'm not sure how that falls into the other categories of the membranes listed.

I'm wondering if commercial RO membranes are intended to let soluble silica pass to avoid premature membrane fouling via scaling. It seems that high pH feed water is the way to prevent scaling if the feed water has a high concentration of silica. I know my feed water has a really high concentration of silicates and I don't raise its pH or have had any issues with premature fouling... possibly a sign that the silicate is not being rejected? I don't know, just speculating :)

I read somewhere a long time ago that RO can't remove any silica at all. It seems thats not the case at all.
 
Most of them in this hobby are DOW Filmtec TFC, which are PA/ CPA (Composite Polyamide) membranes and have a 98 % rejection rate for dissolved reactive silica. The issue with CTA is they need to have chlorinated water run through them as they are organic and foul and break down easily by bacteria.

. Reactive silica is rejected well by thin-film polyamide (PA) membranes, yet is not reject as well by CA membranes

Filmtech

http://www.dow.com/webapps/lit/litorder.asp?filepath=liquidseps/pdfs/noreg/609-00002.pdf&pdf=true
 
Thanks for the info Boomer, I have the same filmtech membrane in my RO as you posted. I might do some water tests today and compare silica in my feed water to effluent from RO vs. after RO and DI. I'm skeptical about my test kit though; I'm using the seachem silicate test kit, but it seems a bit crap to me.
 
I just ran a few silicate tests on my RO/DI water. I tested my feed water, RO water, RODI water, and waste water:

Feed water: 10+ mg/L

RO water: ~0.25 mg/L

RODI water: <0.25mg/L

Waste water: 10+ mg/L

The difference between the RO and DI was extremely minimal and I'm sure the stdev. from testing error would make the difference inconclusive. It's pretty obvious that the RO is certainly removing a very significant amount of silica.
 
No problem,

I'm using a 1 micron sediment filter, followed by two chloramine block carbon stages. The membrane is a DOW FilmTech TFC rated at 75 gpd.
 
check out this thread guys!

http://www.reefcentral.com/forums/showthread.php?s=&postid=11078223#post11078223

they are getting some strange results from AWT!

<a href=showthread.php?s=&postid=11077762#post11077762 target=_blank>Originally posted</a> by SuAsati
Fresh RO/DI water...
------------------
Ammonia (NH3-4) 0.001 Good
Nitrite (NO2) 0.002 Good
Nitrate (NO3) 0.3 Good
Phosphate (PO4) 0.19 Good <----
Silica (SiO2-3) 1.9 High <----
Potassium (K) 0 Low
Calcium (Ca) 0 Low
Boron (B) 0.3 Low
Molybdenum (Mo) 0.0 Good
Strontium (Sr) 0.0 Low
Magnesium (Mg) 29 Low <---- Don't know How I am getting 0TDS?????
Iodine (I¯) 0.00 Low
Copper (Cu++) 0.00 Good
Alkalinity (meq/L) 0.00 Low


Same Water with Tropic Marin at sg 1.023.
----------------------------------------
Ammonia (NH3-4) 0.010 Good
Nitrite (NO2) 0.002 Good
Nitrate (NO3) 0.3 Good
Phosphate (PO4) 0.06 Good <------ This is perplexing???? (see 0.19 in fresh water)
Silica (SiO2-3) 4.2 High
Potassium (K) 355 Good
Calcium (Ca) 387 Good <----- Salifert gave me 410 so not that off...
Boron (B) 1.4 Low
Molybdenum (Mo) 0.0 Good
Strontium (Sr) 8.6 Good
Magnesium (Mg) 1213 Good
Iodine (I¯) 0.01 Low
Copper (Cu++) 0.03 Good
Alkalinity (meq/L) 2.97 Good
 
The 29 ppm Mg might be the error limits of the testing equipment, especially if it's calibrated for saltwater. The phosphate test might also have some calibration issues. Without knowing how the testing is being done, not much can be said.
 
Yeah that is true

i was also thinking maybe their equipment is set up for salt water.

This is only thing they said about it is....

-----------------------------
Do you have a service for freshwater tanks?
At the moment we do not, but we are hard at work developing a truly meaningful panel of test parameters and insights for freshwater enthusiasts. If you are keeping live plants and dosing supplements and fertilizers you should find this service invaluable. Keep an eye on this site for more details.
 
So check me if i am correct in my thinking here
my Pinpoint conversion chart states that 1.023 = 31.1ppt
full strength sea water is 35.0ppt so if one multiplies 31.1 by 1.1255 you will get 35.0 correct?
now go through those test numbers and multiply them by 1.1255 and you get:
------------------------- 31.1ppt ------- 35.0ppt
Ammonia (NH3-4) 0.010 Good
Nitrite (NO2) 0.002 Good
Nitrate (NO3) 0.3 Good
Phosphate (PO4) 0.06 Good ----- .07
Silica (SiO2-3) 2.3 High ------- 2.6 (I subtracted the FW values)
Potassium (K) 355 Good ------- 399
Calcium (Ca) 387 Good ------ 436
Boron (B) 1.4 Low ------- 1.58
Molybdenum (Mo) 0.0 Good --------
Strontium (Sr) 8.6 Good -------9.7
Magnesium (Mg)1213 Good ------- 1365
Iodine (I¯) 0.01 Low ------
Copper (Cu++) 0.03 Good ------
Alkalinity (meq/L) 2.97 Good ------- 3.34 not sure alka will behave like this.

I think some of the phosphate probably precipitated out.

Steve G.
 
Last edited:
my Pinpoint conversion chart states that 1.023 = 31.1ppt

Yes, I know he has that but that is in error. 35 ppt = 1.0264 or 1.023 = 30.5. I don't know where he got that from. We always base it on @25C /77F.

multiply them by 1.1255 and you get

Yes, that is about what you would get if you raised the salinity. However, that 1.255 would now be or should be 35 / 30.5 = 1.1476. So, the Ca++ would be 444 ppm instead of 436 ppm for example.
 
Boomer did AWT ever get back to you about the concerns you had? I been holding off sending in my next sample until they get the problems worked out.
 
Boomer,

Sorry to hear that. I would send in my water for testing, but I'd like a few of these discrepancies clarified. Hopefully, they will respond to you soon.

Joyce
 
well new test results in.....

Water Test Summary
Ammonia (NH3-4) 0.005 Good
Nitrite (NO2) 0.004 Good
Nitrate (NO3) 0.3 Good
Phosphate (PO4) 0.04 Good
Silica (SiO2-3) 1.2 High
Potassium (K) 295 Low
Calcium (Ca) 357 Good
Boron (B) 4.3 Good
Molybdenum (Mo) 0.0 Good
Strontium (Sr) 9.9 Good
Magnesium (Mg) 1188 Good
Iodine (I¯) 0.01 Low
Copper (Cu++) 0.02 Good
Alkalinity (meq/L) 3.21 Good

I got some changes to make.. :(

What would be a good plan?
 
There is nothing to really change there other than the Ca++ and Mg++.

As I have said NSW silica is NOT as low as they claim, it is 3 ppm. We recommend not to go above 1 ppm and you are at 1.2 ppm. There is not need to add iodine and IMHO Potassium either.

It is really funny that they say 0 ppm Mo, Good, yet is is required but all photosynthetic organisms. And no I don't think you need to add any of that either.

And even funnier is they say your Ca++ and Mg++ are , Good, when both are really low :lol: There are no places in the open ocean or on any reef where you would find those levels. That would equal a salinity of 30 ppt.


What salt are you using USC ? Right now you would to need to add ample amounts of Calcium Chloride and Magnesium/Sulfate/Chloride.

And all this is based on if you want to believe their test results.
 
I been using reef crystals. I believe the reasons for my off MG and ca is that i had an kalk overdose. Nothing too bad, ph never got over 8.6 but i'm sure this is the reason for the off results.

I went today and brought all new test kits[API & salifert] my old test kit tested CA++ at 400 but now i know it was way off. Salifert tested CA @ 340.. :( MG @1150.. :( Alk @ 10dkh

I brought an iodine test kit/supplement and also some seachem flourish potassium.

It would take A LOT of the potassium supplement to raise it. I will just dose 15ml weekly. This is raise it 3ppm per week.

At least my P04 and silica are way down. :)

Thanks for the help!
 
Back
Top