A variation in floc size means there is a change in the quality of the effluent (or sludge) or insufficient reagent dosing/degraded quality (aging polymers, for example0.
-In the case of effluent, first check the reaction pH and adjust if necessary.
-Then, in all cases (whether sludge undergoes flocculation or effluent receives physical treatment), do the following:
-Ensure reagents are properly injected.
-Check treatment concentrations (i.e., that the initially prescribed dose of reagent is properly injected)
-Check the quality of the reagents. Watch out: prepared polymers are not stable over time (should not be stored for longer than one day in the case of cationic polymers, which hydrolyze quickly; and in the case of anionics, no more than a few days).
-Redo (or do) jar test to determine the optimum treatment rate if all of the previous tests are OK or if the quality of the effluent (sludge) has changed.
Of course, it goes without saying you have to adjust the settings in the field based on the observations and lab tests. And if the polymer is old, do it again...
My two cents is that in addition to the problem of coagulant and polymer treatment concentrations and checking the coagulant dosing and pH levels, you also want to check contact time in the coagulation tank. When the temperature of the water drops (and it's the season...) contact time for coagulation is greater. This can be determined by jar testing: apply the optimal dose of coagulant and flocculant and vary contact time. Certain coagulants also less effective in cold water, so you may have to switch to a different one in winter.