DIY RODI

[RocketEngineer]

That membrane is not going to work as you intend for two main reasons: 1) not enough pressure, and 2) comparatively low rejection. Based on what I can find, this membrane needs 200psi to achieve 95%. How much water do you expect to filter at once? I recommend you look into a boost pump to add after the filters but before the membrane.

 

[WillpoleReefers]

That membrane is not going to work as you intend for two main reasons: 1) not enough pressure, and 2) comparatively low rejection. Based on what I can find, this membrane needs 200psi to achieve 95%. How much water do you expect to filter at once? I recommend you look into a boost pump to add after the filters but before the membrane.

Thanks, I am definitely going the booster pump route

Steve

 

[vahegan]

Yea I thought it might be necessary, was curious to run it without initially and see what was possible though. It’s not a big issue to add a pump if needed,

Steve

That membrane is not going to work as you intend for two main reasons: 1) not enough pressure, and 2) comparatively low rejection. Based on what I can find, this membrane needs 200psi to achieve 95%. How much water do you expect to filter at once? I recommend you look into a boost pump to add after the filters but before the membrane.

Yes, was going to write exactly the same. Rejection rate of the membrane is much higher with higher pressures, and I believe a booster pump is a must.

As for the membrane itself, I'd rather go with Dow Filmtec (now owned by Dupont), they make the best membranes. I have TW30-3012-500 which is for 500gpd.

I just checked, it seems that currently it has been replaced by Axeon TF-3012-500. I guess, there was another change of the brand recently... They are not exactly cheap, something in the $140-150 range for the 500GPD model. I had tried some unknown brand 400GPD membrane for $50 on Amazon, and that was complete crap, had to ditch it immediately.

P.S. Note to check the manufacturer datasheet for the membrane about maximum pressure. For the above mentioned membrane Dow says up to 10 bar (150 PSI). I am using a Shurflo 8000-553-890 pump which gives up to 80 PSI.

P.P.S. Here's a good PDF from Dow, about how different factors affect rejection and flow:
Factors Affecting RO Membrane Performance

 

[WillpoleReefers]

Yes, was going to write exactly the same. Rejection rate of the membrane is much higher with higher pressures, and I believe a booster pump is a must.

As for the membrane itself, I'd rather go with Dow Filmtec (now owned by Dupont), they make the best membranes. I have TW30-3012-500 which is for 500gpd.

I just checked, it seems that currently it has been replaced by Axeon TF-3012-500. I guess, there was another change of the brand recently... They are not exactly cheap, something in the $140-150 range for the 500GPD model. I had tried some unknown brand 400GPD membrane for $50 on Amazon, and that was complete crap, had to ditch it immediately.

P.S. Note to check the manufacturer datasheet for the membrane about maximum pressure. For the above mentioned membrane Dow says up to 10 bar (150 PSI). I am using a Shurflo 8000-553-890 pump which gives up to 80 PSI.

P.P.S. Here's a good PDF from Dow, about how different factors affect rejection and flow:
Factors Affecting RO Membrane Performance

Thanks so much for all that, particularly the Dow document. Now saved here. They say RO can be a complicated business, no ****. I now understand why pressure is so important … constant salt leakage vs pressure dependent reverse osmosis flow. I think the next step with my unit is to add a pump. Measure whatever rejection ratio I can achieve across the membrane I have, which based on what you say may be sub optimal as an eBay sourced item, then consider whether to stay with it. I have not ruled out a much bigger deionisation arrangement, possibly separate anion/cation

Steve

 

[KStatefan]

If you really want to take a deep dive into membranes here is the Filmtec Manual.

Before you make any changes you should look at the specs of your current items and to me more important the quality of your source water. That is one of advantages of putting together a RODI system you can build it to your needs.

 

[WillpoleReefers]

If you really want to take a deep dive into membranes here is the Filmtec Manual.

Before you make any changes you should look at the specs of your current items and to me more important the quality of your source water. That is one of advantages of putting together a RODI system you can build it to your needs.

Another useful .pdf thanks!

 

[WillpoleReefers]

Just wondering if anyone following here has used one of these commonly available (eBay, Amazon etc) conductivity interface boards. Outputs 0-2.3V for 0-1000ppm nominally. Had my first play and measured the raw water here at about 240ppm un-calibrated. Wonder how accurate they are un calibrated vs with a calibration point or multipoint curve. Easy to input 4 sample points into a micro this way, no custom analogue stuff to fabricate.

Steve

 

[Woodyman]

Just wondering if anyone following here has used one of these commonly available (eBay, Amazon etc) conductivity interface boards. Outputs 0-2.3V for 0-1000ppm nominally. Had my first play and measured the raw water here at about 240ppm un-calibrated. Wonder how accurate they are un calibrated vs with a calibration point or multipoint curve. Easy to input 4 sample points into a micro this way, no custom analogue stuff to fabricate.

Steve

Never used one like that. I've used the prebuilt variety. Which for aquarium purposes is overkill.

 

[WillpoleReefers]

I guess the point of accuracy is picking up early deterioration in output quality after the membrane or RO DI stages so that relevant replacements can be done. I wonder what long term drift will be like on output from such sensors. Be interesting to see. I’m intererested in comparisons over time, a micro with multiple inputs is well-placed to do this, monitoring rejection rate, resin performance etc. I do know I need temperature measurement and correction to achieve best accuracy, that is in the plan,

Steve

 

[WillpoleReefers]

Forgive me commenting my own post, but I now have some data if anyone is still reading this, it is interesting . I made a stack of three of the little conductivity boards, pictured below. I then sampled conductivity at three points, raw feed, post membrane and output. Initial data was pretty haywire, I ended up thinking my resin was exhausted, turns out one of the three boards was faulty, but it gets confusing when you are holding probes on to measure manually and there are 3 inputs. Turns out the faulty board was outputting about 0.4V whatever the conductivity. Discarding that boards data and swapping inputs to switch them the other two boards gave very consistent results. Anyway both resin cartridges were refilled with fresh! The data is below, excuse the lack of precision, I also managed to wet my better Fluke meter in the process today, took it out for a while. TDS is calculated using volts/2.3 x 1000 (full scale TDS) Anyway:

Output volts/ TDS
Raw 0.62 / 269 ppm
Post Membrane 0.025 / 10.9 ppm
Post Resin (X2) 0.002 / 0.87 ppm

Rejection ratio of membrane = 269-10.9/269 = 95.9%

So it would appear, if the data is good, that running a 400gpd membrane at 120 gpd and 42psi is not as bad as might be expected. I guess the rejection ratio will improve from here if I add a pump.

Morals learned today. If using cheap boards for conductivity test on known samples before installation! Secondly don’t leave Fluke meters anywhere near an RO setup you are working on,

Steve

 

[WillpoleReefers]

Getting numbers out now. Before and after. Uncalibrated!