If you speak the plant cell, it is the internal concentration of the cell that determines the degrees of salinity reverse osmosis.

Dear Ruediger, You may find the answer in this interesting document.

JC

https://www.usbr.gov/research/AWT/reportpdfs/report119.pdf

Dear Abdelhakim, it is the Salinity of the feed water to the RO.

The salinity  is reduced by 95 to 99 % for each pass through a membrane . The product water quality  ( TDS)  can be practically  nil ( semiconductor industry need ) by selecting the numerous of passes on product side.The salt in the product water is passed onto reject water.Each pass on reject side can yield  about 50 % of feed water. One can recover almost 8 to 90 % of feed water. THE RO Becomes uneconomical for reject water quality beyond about 10 % salts (100,000 mg/l).Beyond this one has to use thermal evaporation to improve the yield  or recovery. All in all for sea water normally product TDS is  100 mg/l with 85 % product yield. For brackish water product TDS of 30-50 mg/l recovery 90-95%. Is practical. Beyond this process becomes expensive to operate. 

Any RO membrane manufacturer can give you  optimised solution based on needs and feed water quality.

Parkash has explained very well the RO during desalination of saline water. There is no limit of salts fixed for feed water. In Gulf countries, the desalinated water is mostly produced from seawater or groundwater of almost similar quality with a few exceptions. However, there are three things very clear in this regard.

1. Higher the salt content of the feed water lesser will be the produced desalinated water and more will be the brine water (Reject).

2. Higher the salt content of feed water, more number of membranes have to be added in the cycle.

3. Higher the salts in the feed water, higher will be cost of desalination and the system will be becoming lesser and lesser economical

Hence, for an efficient, effective and economical system, lesser saline water is preferable but depending upon availability. 

Dear Nasir,

Thank you for your detailed explanation. - In my case we have a salinity (dissolved salts) in a range between 150,000 and 200,000 mg/l. That is a range, where RO reaches their limits, what you also have mentioned. I have learned in the meanwhile, that NF would be more capable for that problem!?

Dear Ruediger, Thank you for your comments. You are right, when the salt content in feed water are too high, the engineers recommend Nano Filtration (NF) to reduce EC to bring within limits where RO system start working. The pressure used to run the system is also very important because it controls the flux.

Nice question  Ruediger . I agree with Dr Nazir , I don't think, there could be any maximum limit of salinity of the feed water, since Ro technology ois latest of all the available options for desalinization .however the maximisation of recovery  as the ratio of treated water  produced to source water used  is the major challenge in such technology, besides an effective  disposal  of saline concentrate produced as RO product . The recovery based on salt saturation indices derived from feed water quality serves a better purpose. The only disadvantage in Ro technology is that it needs massive pre-conditioning of feed water so that  shelf life of Ro membrane is extended. Find below two PDFs , could be of some use.

Thank you Anoop for your kind input and appreciation.

Interesting question Ruediger Burkhard Richter .. i think for 150,000 mg/L salt, you might better use thermal desalination instead of membrane desalination - that is likely to be more cost effective.