Normalizing performance ...

Normalizing performance relative to the manufacturer's specification of test data for the membrane model number installed (can be still useful in a mixed system if set up correctly), tells the operator several things:  the corrected net driving pressure at specified flow rate (hence also the normalized permeate flow rate at standardized pressure, as corrected for osmotic pressure, temperature, and average membrane applied pressure. 

The normalization program then also provides the normalized salt rejection (and/or salt passage), which again is referenced to standard conditions of temperature, and corrected for recovery at the average membrane. 

The differential pressure along the length of a vessel containing a number of membranes (often 6 elements), is also calculated from entered data, and a standardized differential pressure is calculated and this can be one of the key indicators of serious fouling within the spacer between leaflets of membrane material.

Some normalization programs may have chemistry entries for sparingly soluble species such as Calcium (with carbonate or sulfate), and also for Silica.  Unless these calculations can also provide some indication of concentration gradient at and near the membrane surface, they can miss scaling indications.

Most cases, a net loss of 15% of permeate flow is a strong indicator of scaling/fouling, and the fore-warned operator will begin a CIP (clean-in-place) sequence, alternatively selected membranes can be removed from the system and cleaned even individually, although this becomes very labor intensive.  If differential pressure coefficient and/or salt passage reveal statistically significant increases ~20%, cleaning is also recommended.  Caveat (warning): cleaning of an entire vessel or parallel set of vessels simultaneously may result in only shifting the fouling  or scale down the chain of elements, with a very ineffective result, where performance appears worse after cleaning.   In my opinion, this is usually the result of attempting to clean with too low a linear flow velocity along the length of the elements.

In some cases, cleaning may be done with flow velocity counter to the normal feed direction to attempt to flush out particulates that are partially clogging the spaces between membrane sheets (envelopes).  Another technique is the very careful application of reverse pressure application from the permeate side of the membrane, although if not done with extreme control, can result in ripping the glued seams of the envelopes opening, and destroying the effective salt rejection of the system.  It is an effective way to induce sales of new membranes, however, as unethical as it is.