saturation in simulation

[tinni]

hello everybody,
i have one question regarding the basic concept of water saturation and its use in reservoir simulation. in literature we deal with two types of water saturation. connate water saturation and irreducible water saturation. connate water saturation is developed during the initial phase of migration. if connate water saturation is less than irreducible saturation, water is not movable. (please correct me if i am wrong).
now during my gas reservoir simulation, i only have SWL (connate water saturation) and relative permeability SGWFN data. i set a value of 0.2 and .1 for my critical water saturation SWCR and critical gas saturation(SGFR) values respectively. i set a maximum connate water saturation value to .8 . now when i run my simulation its showing error messages like this

CONSISTENCY PROBLEM WITH WATER PHASE END POINTS IN GRID CELL ( 6, 32, 1 ) CONNATE= 0.8000 CRIT= 0.2000 MAX= 1.0000

but if i copy SWL for SWCR, its running without error. now i am confused. is it something like that the SWL values are treated as irreducible values? observed production data show that water is produced with gas from the very beginning of gas production. if SWL and SWCR values are same, how can i get water production from the beginning? should i use any keywords like SWAT or SWATINIT to define initial saturations? should these data be diffrerent from SWL data? and if so, what to do as i dont have any other data?

please help me to solve my problem

[catapam]

Regarding what u say I have following comments:

Smin - is saturation minimal (this is what has to be regardles how we produce the field except heating or surfactans are used)
Scr - is saturation bellow the fluid won't flow , once over critical the fluid will flow

conate saturation is something that originally is in reservoir.
i think your connate water as 0.8 is too high

at 0.2 crit connate should be bellow 0,5

i'm also beginner in simulation and what i generally used is SWOF, SGOF without SWCR, SWL
I think SWCR, SWL is anyhow taken from rel perm curves

I could be wrong but this is how I understand