Reliable RO plant for high silica groundwater source 4

[DoctorSludge]

We have several groundwater borewells with up to 42mg/l SiO2, total mineralisation of about 4000mg/l but not much iron or manganese to worry about. The water obviously has a tendency to 'scale'. Our existing Reverse Osmosis plant needs three people to look after it 24/7, was built cheap and fast and fails regularly. When it fails, the consumers use more detergent and push up the phosphorus in the final effluent to the detriment of the consent to discharge. I am looking for an automatic, packaged, plug and play, reliable, 'build like a brick' tried and tested RO plant that will work for 20 years, inform the control building when it fails and needs the absolute minimum of chemical addition and an annual inspection and parts replacement. We have plenty of water, could waste some for backwashing and need a production rate of 18000 gallons per day (we will probably use 6000 gallons per day (27 cubic meters). Can you recommend a manufacturer and or supply experience that could help us to do this. Where are the Mercedes Benz and Rolls Royce RO plants and do you have one we can talk about? Thanks.

 

[Shaman2]

Could you supply more info on your current system? How many stages? How many membranes per stage in what configuration? What concentration factor at each stage? What type of membranes? What do you mean by "cheap and fast", undersized so you are trying to push more water through it than it's design can handle?
Regardless of design if you increase the concentrations of minerals retained too high they are going to scale out. More cross-flow can help sweep them downstream as solids, but unless that is in your final stage it is just adding to troubles downstream. The lower you can keep your concentration factors the better.
Does your set-up allow for individual sampling of the tubes of membranes to isolate any patterns to the fouling?
Your comments about consumers using detergent leads me to the conclusion we are considering domestic water usage, so how clean are you trying to go with it?

 

[quark]

Generally, membrane life is about 5 years and you do have to carry out chemical cleaning once in a year. There is no such system as trouble free RO if your pretreatment is a mess. Magnesium forms very hard scale with silica (dissolved) and chances are that the membrane pores will be plugged forever.

You have to either go with clariflocculation or UF as a pretreatment for RO. 27 cu.m/day is rather small.

 

[bimr]

If you buy your system from the major water treatment equipment suppliers, there is probably not going to be much of a difference in reliability. Contact US Filter.

If you buy from some of the smaller garage type organizations, then it is buyer beware.

You will get what you pay for.

 

[stanier]

Silicon is the killer for membrane plants. Always has been. It is extremely difficult to remove during a cleaning process.

Contact Koch and talk to them about it.

Geoffrey D Stone FIMechE C.Eng;FIEust CP Eng

http://www.waterhammer.bigblog.com.au

 

[Muggle]

If you can't find the RO process/membrane you want, you might look at electrolytic oxidation to remove the silica/silicates and other minerals before the RO. We ran some tests for National Gypsum - reduced the load from municipal reclaim water by 75% before the R/O. It's simple, rugged, and does not require daily or weekly maintenance - no high pressure - no chemicals, etc. Your levels are really low - normally, we deal in several 1000' ppms.

Yes - the R/O and UF guys are going to "poo poo" this suggestion -

Dave/Aquatic Technologies

http://www.eoh2o.com

 

[DoctorSludge]

Thank you everyone for the help. Some more details. The existing system comprises a set of primary filters - cartridge type then a multimedia filter. The Membrane appears to be a single pass membrane but surrounding this package is a set of chemical pumps. These provide pre membrane chlorination, dechlorination, post membrane chlorination and the addition of descaling chemicals. At the end of this process we are looking for domestic water that allows soap to make a lather, is microbiologically safe and meets the EU standards for potable water. The input sulphate level is high at 2400mg/l and this is the primary reason for the unit - we have a goal of less than 200 for this ion. We would be happy to run with a system that used more water - we have a hundred cubic meters per day available - and we will need 27 for use. The magnesium and silica scale could be a problem. We have no chemical storage facilities on site and health and safety is a major issue so I am inclined away from chemical additions outside the annual service and minded to engineer for dilute concentrations, filters that we can monitor via pressure drop sensors and the electrolytic oxidation is worth a look. I will check out US Filter - thanks guys. If you know of any great installations with managers I could talk to I would like to do that but we are progressing.

 

[sbush]

As silica and total dissolved solids (TDS) levels rise, so does the frequency and magnitude of maintenance in RO systems.

RO is not always the best method for demineralizing water in every case. Many municipalities with high silica and other high fouling/scaling water supplies have either replaced or added EDR to their RO systems. If you are interested, try contacting Ionics, Inc (USA). They have years of experience with EDR systems.

S. Bush

http://www.water-eg.com

 

[DoctorSludge]

We have a cationic and anionic resin column in the cabin for ultra pure water for industrial use, however, we are in a remote location (a days flight from anywhere industrial) and the health and safety issues of recharging the columns with strong acid/alkali rules out future use.
We are still looking for the golden filter!

 

[dfmorvan2]

Hi Dr Sludge

42 mg/l is pretty high for silica but all depends on the concentration ratio of the RO and its arrangment like shaman2 said. Silica solubility is around 100 mg/l at "usual" pHs (around 7-7.5). It is higher at low pHs and much higher at high pHs. It is indeed very difficult to remove like quark and stanier wrote.

I have experience in cleaning and running a RO plant with 70 mg/l SiO2, 2500 mg/l Sulfate and 1000 ppm CaCO3 Hardness. You can use several types of antiscalants, some can work up to 160 mg/l SiO2 in the concentrate others up to 200-220 mg/l. After this everything depends on follow up on the unit so to detect early scale precipitation as Silica is very hard to remove (You will need at least 20 hours cleaning the membranes in ph 11 solution+EDTA+dispersant to begin to remove something).

So to summarize :

- get a full water analysis with Ph taken on site. Do not forget to ask for Barium and Strontium as they combine with Sulfates.
- find a reliable supplier of membrane antiscalant who has some software to predict antiscalant performance
- let him do the calc with his software and the analysis
- Make him check with the membrane manufacturer software that the conditions he gives you (conversion ratio mainly) is OK with the hydraulics of the membrane plant.
- Get a normalization program from a main membrane manufacturer website and begin logging every day the required data from your RO (Pressure, conductivity, temp, etc..)
- As soon as you see a shift of 10 % in the normalized flowrate or salt passage it's time to clean the RO. You can see with your chemicals supplier or try to make your own washing solutions.
- Check if your cleaning procedure is OK by going on membrane manufacturers sites (or contact me if you need links).

For your information the plant I'm speaking of has 70 mg/l silica a little iron, Hardness is 1000 ppm CaCO3, TDS around 4000 ppm, sulfate is around 2500 mg/l. The customer cleans every 1-2 Months. Cleaning takes one full day (acid and caustic).Conversion ratio is 50%.
Flowrate is 50 m3/h @ 14°C

Membranes lasted 5 years.

Regards - DF Morvan

 

[stanier]

If you have energy to burn you could look at flash evaporisation rather than RO. If you are in a remote location I guess you generate your ow electricity. It may be feasible to run the generators overnight to treat enough water for the next day?

Geoffrey D Stone FIMechE C.Eng;FIEust CP Eng

http://www.waterhammer.bigblog.com.au

 

[cub3bead]

You will still have the problem with amorphous silica precipitating only it will be on the tubes rather than the membrane surface if you are not careful. And it is not only extra money for the energy, the capital cost for an evaporator maybe 2X a membrane plant.

 

[sbush]

EDR (electro-dialysis reversal) does not involve ion exchange resins. Even though the name sounds like ion exchange, Ionics is primarily a manufacturer of EDR systems. Check them out at

http://www.ionics.com.

They might have the silver bullet you are looking for.

S. Bush

http://www.water-eg.com