kby
Full Member
Posts: 162
|
Post by kby on May 19, 2008 19:31:52 GMT
So, my RO unit appears to have failed this weekend (just in time for summer & the heat wave we have been having). I thought it was the membrane, but, of course, after having spent the $100 on filters and putting them in, it appears that's not the case. Symptom is it doesn't shut off, as well as the water is tap. It started slightly worse than tap before cycling through once, but that is probably due to the fact that the membrane preservative is still being washed out. Initially the symptoms were rapidly rising TDS then noticed it hadn't shut off in 8+ hours (normally about 2 to fill the tank then it shuts off).
This unit is GE, but I think it's a re-branded Ametek. The previous one I had was also a GE; quite similar unit.
However, they've each lasted less long than I would have thought—about 2-3 years. The consumables usually last longer than expected, even though our water's pretty hard, but I expected that to be mostly a consumables (filters & membrane) issue. The inlet pressure is probably on the high side but I don't think it is excessively so.
Since it's kind of an emergency I'll probably go ahead and get the same one again just so I can just swap the hoses around to install it (undersink unit). However, I'm not happy, so I have a couple of questions:
1. Is this type of failure common and what causes it (I'm guessing some internal seal is leaking)?
2. Is it fixable in general or on certain models?
3. What's the usual expected lifetime for the unit? Is this average or short?
Thanks!-kby
|
|
|
Post by Not a Number on May 19, 2008 19:47:22 GMT
Butch (av8tor1) is the RO unit expert. I would check to see that the "O" ring seals are in place on the filters and that there are no old "O" ring or other debris inside the unit where the rings would seat.
|
|
|
Post by av8tor1 on May 19, 2008 20:22:41 GMT
Kby, Warren is too kind, I dont know about the expert part but I will try to help if I can. Does the system use a captive air tank with a hydraulic ASO valve?.... If so, the pressure downstream of the valve is not reaching 60% of the inlet pressure or the valve is bad. This would cause the unit to not shut off... but taken with the TDS readings, this is probably a symptom of the problem and not the root cause. It sounds as though the inlet pressure to the unit has dropped to a critically low level. Do you have a pressure gauge for troubleshooting purposes? Most saddle valves use a rubber gasket, sometimes these gaskets wil begin to restrict water flow, resulting in a large pressure drop and cause the condition you are describing. You might try screwing the piercing valve in and out a few times and see if that opens up the restriction. With low pressure the brine rate will be excessive and the unit will run for hours without ever building up the required 60% outlet pressure needed to activate the ASO valve a pressure gauge and TDS meter will be required for proper troubleshooting, without them I can only make educated guesses HTH's Av see these links for more info: terraforums.com/forums/showthread.php?t=109118terraforums.com/forums/showthread.php?t=112605&highlight=reverse+osmosis
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 19, 2008 20:30:29 GMT
av8tor, Thanks for the hints; I'll check them when I get home. I think however this one has a tap that screwed in. But it could be clogged as well. I don't have a pressure gauge, unfortunately, but will at least see if I can see anything obvious. I would have expected a leaking rubber seal at the saddle valve (if there was one) to show leakage?-kby
|
|
|
Post by av8tor1 on May 19, 2008 20:34:01 GMT
Kby, the condition i describe happens pretty frequently... with low inlet pressure the unit will run continously and the output TDS will climb due to the resulting drop in rejection ratio
look for some type of flow issue/pressure drop to the unit
Av
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 19, 2008 21:33:46 GMT
Thanks; if it's that it will be a lot cheaper to fix :-). Will look tonight & report back.
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 20, 2008 17:49:50 GMT
OK, here's what I know. As I said, I don't have a pressure meter (unfortunately), but I did pull off the connection going into the pre-RO filter cartridge and the flow seemed quite good and fine (about what was expected). I also then pulled the connection that goes into the RO membrane housing, and it too showed a higher pressure (but smaller) stream. I then noticed when draining the system (figuring I'd have to go ahead and replace it, which is what I ended up doing, but I'm interested in repairing or at least understanding the failure mode) the storage tank didn't empty (even after no flow at the faucet when the inlet is shut off there's a fair amount of water you can hear sloshing around inside). The last unit I had also seemed to exhibit this behaviour, although I don't remember it being in continual run mode (but might not have not have caught it as I did not catch the poor water quality until way too late...oops). I do remember it showed a symptom more readily tied to the storage tank, namely, it didn't seem to store as much water. I've tried repressurizing the tank on that one but may not have done it enough. I haven't noticed at least as drastic a reduction in storage capacity with this one, at least not yet, but that doesn't mean it's not reduced from what it was when new.
As I alluded to above, in the interests of time (at the unfortunate expense of $$$) I decided to just replace the unit and the new one is behaving more normally (down to about 30ppm; about 19 is the best I've seen before and that can be seasonally dependent due to the water mix at the faucet). But I'd still like to know what the failure is to either avoid or stock up on spares (e.g. if it's the tank it can be ordered as a part but not necessarily fast in an emergency). Thanks!-kby
|
|
|
Post by av8tor1 on May 20, 2008 19:01:00 GMT
Seems as though you are describing a ruptured bladder, but I don't see how this would result in the unit running for hours without shutting off, in fact Iwould expect it to shut off faster due to the fact that you no longer are compressing the precharge
Without actual pressure readings, I don't know what to tell you at this point. These symptoms see to contradict each other, very odd indeed.
I'm curious... what is the PPM without the tank in the circuit... does it return to normal?
and the normal precharge on a bladder tank is about 8 psig
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 20, 2008 20:47:01 GMT
I did check the output to the tank in the old unit before and it looked like about tap. I thought this might be due to not completely flushing out things, but I've noticed (with the new unit) that it flushes out the preservative (or at least it doesn't register on the TDS meter) pretty fast (even though the instructions say to fill & empty the tank 3 times). Could be more than one problem, of course, but is the bladder tank something that dies a lot? I think the new one (same as the one I just replaced) had a 14-15PSI normal precharge label (and a big warning not to mess with the charge unless you were a professional). The original one had no label. On that original one I think it is more clear that there was a problem with the internal membrane as water came out the charge valve when the needle (normal schrader valve) was depressed. I haven't fiddled with the one I just replaced yet. I do think you are right about the original one in that it shut off too soon as a symptom of the tank having a problem; that's why it only had a tiny amount (of not very good quality) water in it.
Can you give me a Cliff notes version of what's in the bridge that connects things from the carbon filters to the RO unit (e.g. from the pre-filter cartridge to the RO chamber)? Are the seals that can die or is it simply molded plastic passages? Thanks again for all the help; I'll see if the one I just replace has water in the pressurized part or it's just not empying enough.
It seems on the original one that when I drained the water out (via the schrader valve) then pressurized it "some" it didn't seem to leak into there any more which also doesn't make sense. It's attached to an outside faucet, but seems to at least shut down some (maybe not completely) but the quality isn't too good (but I haven't really tried to troubleshoot it, either).
What is the normal minimum pressure required and what should it be at at the RO membrane input (should I hever get a meter)?-kby
|
|
|
Post by av8tor1 on May 21, 2008 1:25:02 GMT
Kby, from the prefilter stages to the membrane is pretty much just plastic passages and an ASO valve... nothing really there to fail. the water leaves the prefilters, passes through the ASO valve and then flows across the membrane surface, part of the flow is then allowed to exit via the restrictor orifice to the brine output. This restrictor is sized based upon the membrane flow rating and an assumption that inlet pressure will be somewhere between 50-60psi, Based upon these values enough pressure is being applied to the membrane to give you a rejection ratio in the high 90 percentile range. At the same time a brine rate of about 3:1 will be maintained. This is what keeps the membrane surface flushed clean of debris and gives it the long life. In the permeate outlet from the membrane you will find a check valve which prevents reverse flow through the membrane. This would occur anytime there is any drop in inlet pressure (someone flushed the toliet...etc) then from there it goes into the control side of the ASO valve, then on to the DI stage and/or bladder tank. Once the control side pressure of the ASO valve reaches 60% of the membrane inlet side, the ASO shuts off flow to the inlet of the membrane... if inlet pressure is low, the majority of flow will go out the restrictor and not through the membrane, so the unit continues to run... at the same time rejection ratio drops due to the lack of pressure differential (I know, seems counter intuitive) so the unit runs excessively and TDS climbs... now as far as repeated bladder failure, I dont know what to tell you... this is not a common problem in my experience. These same captive air tanks are used on most homes with wells... they just dont rupture very often. Possibly running the unit completely dry may stress the bladder... but I run mine all the way empty and never have had a bladder failure. I just dont know what to tell you there... if it happened more then once I would look for a different source for my next tank though. As far as membrane life, a great deal depends on brine ratio, usage rates, inlet TDS, rejection ratio, chlorine etc. Check your brine ratio if possible... then go from there In the future I would recommend getting a cheap pressure gauge... it and a TDS meter really take all the voodoo out of troubleshooting RO systems some of my graphics on the subject Edit: adding additional comments as I think of them LOL * bladder tank must be isolated from system during brine ratio measurements * hydraulically operated automatic shut off valves will typically shut off the inlet to the membrane when the outlet pressure reaches 60% of inlet pressure * In a real world system getting 85% of rated flow is about the maximum attainable** * at 85% a 100gpd system should produce about 7.5oz/minute (bladder isolated) * In bladder tank systems, membrane efficiency drops proportionally as the bladder tank fills. This is due to backpressure on the membrane outlet.*** ** inlet pressure, water temperature and TDS will have a major effect on membrane output (TFC membranes are usually rated at 77 degrees Fahrenheit @60psig and for every 1 degree drop, production drops 1.4% (57f water would reduce efficiency by 28%)) *** This drop in efficiency can be prevented by the use of a permeate pump and 90% automatic shut off valve. The permeate pump isolates the membrane from this build up in back pressure. Since the pressure differential across the membrane now remains constant, membrane efficiency also remains constant resulting in greatly improved recovery rates. This isolation also allows us to replace the 60% ASO valve with a 90% ASO thereby utilizing more of the bladder capacity and increasing output pressure. These modifications will result in a 60-80% drop in water consumption and brine output. In addition, this drop in water consumption also results in a proportional increase in prefilter life. A nice, simple and fairly inexpensive mod for any system using a bladder accumulator tank. my home system: RO/DI/PP/90% ASO system: blue = permeate blk = prefiltered water yellow = brine red = fresh water inlet white = bladder tank clear = system inlet and outlets HTh's Av
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 21, 2008 3:49:04 GMT
Thanks, AV, that helps a lot. This one isn't so fance (only one carbon prefilter, no demineralization beads but I get a much better idea of what's where, and know what's expected to fail (or not). Much appreciated!-kby
|
|
|
Post by av8tor1 on May 21, 2008 13:21:51 GMT
Your welcome Kby, most problems associated with RO systems can be traced back to pressure as being the root cause... anytime a system is acting strange a pressure check at the membrane inlet is usually the initial test... it should be at least 40psig, if not then it simply will not work right, brine rates, TDS and run times will be excessive... If the pressure is low then check the inlet pressure to the prefilter stage... if it is ok then replace the prefilters, if it is low then inspect the saddle valve for obstruction... if inlet pressure is ok to the membrane then check rejection ratio, if the pressures are good but rejection ratio is bad... replace (may try a manual flush first) the membrane
(always check pressures dynamically, not statically... deadhead pressure readings are worthless, have the system outlet open and water flowing)
simple and by the numbers, but requires a gauge to do it....
Av
|
|
|
Post by av8tor1 on May 23, 2008 17:59:36 GMT
just for reference, I noticed in another thread there is some confusion about RO vs. DI vs. RO+DI
Here is what little I know about the subject...please correct any false statements, I'm definitely not a chemist.
Deionization and Demineralization by Ionization both refer to specialized forms of Ion Exchange. (both terms are often used interchangeably, but there is a slight difference...true DI will also remove bacteria)
Deionization removes ions as does Ion Exchange. Ion exchange deionization (DI) columns use synthetic resins similar to those used in water softeners.
Deionization systems use a two-stage process to remove virtually all ionic material remaining in water. That is why we call the process Deionization.
In Deionization two types of synthetic resins are used, one to remove positively charged ions (cations) and another to remove negatively charged ions (anions). Cation deionization (DI) resins ions remove cations, such as calcium, magnesium and sodium and replace them with the hydrogen (H+) ion. Anion deionization resins remove anions, such as chloride, sulfate and bicarbonate and replace them with the hydroxide (OH-) ion.
In Deionization, these displaced H+ and OH- ions combine to form additional H2O
In Mixed-Bed deionization systems, the anion and cation resins are blended into a single tank or vessel. Generally, mixed-bed systems will produce higher quality water with a lower total capacity than two-bed systems. Mixed bed deionization resins have design capacities and must be typically be replaced upon exhaustion. To extend bead life this stage is usually performed after the water has gone through the RO process.
Deionization can produce extremely high-quality water in terms of dissolved ions or minerals, up to the maximum resistance of 18.3 megohms/cm. RO/DI are the standard for the pharmaceutical, Electronic and Nuclear industries.
HTH's Av
(and please forgive my copying and pasting, just wanted to be relatively sure of my facts)
|
|
kby
Full Member
Posts: 162
|
Post by kby on May 23, 2008 18:44:21 GMT
I think you've got it mostly right, at least as far as I understand it. I am not sure if you can regenerate ion exchange resins or not, or, even if so, if it's economical for the home user.
I am not sure about the DI vs. DM distinction you mention; that might be kinda fuzzy. Ion exchange in and of itself would only do this as a side effect; that is largely a filtration issue which can be done by any one of a number of filtration type techniques, including RO. So if "true DI" includes a filtration phase of some sort (including RO) ahead of the ion exchange component that might be what is "qualifying" it as removing pathogens. Pathogens should be larger than any ions being exchanged by the resin, which basically work by having molecular cavities (and charges) that have an affinity for the particular things you're trying to remove.
Strictly speaking "deionization" isn't a correct term as there are always H+ and OH- ions in water, just and a small and well defined concentration (defined by the molecular properties of water itself). "ion exchange" is a better term.
3x distilled from potassium permanganate in the first stage (iirc) was (even though ion exchange & RO were around) was the standard in an electrochemical research lab I worked in in the mid 1970s--much more ion sensitive than any CP. I'm not sure what the current standard is. That technique would not scale well economically or in terms of maintenence, even for the casual grower.-kby
|
|
|
Post by av8tor1 on May 23, 2008 21:14:41 GMT
It is my understanding that single bed media may be regenerated, but I would think that mixed bed would not be feasable due to the conflicting regeneration requirements I totally agree in the incorrectness of "Deionization" excellent reply Kby, thank you... it definitely encourages more discussion on the topic, well done! Butch
|
|