Issue with Volume of Peripheral Comp

Dear All, I am having a strange issue with my model. I have a dataset after oral administration and samples are collected on time points 0,1,2,4 and 8 hours at steady state. I fit this data with ADVAN5 to a 2-comp model. I am getting volume of peripheral compartment estimated as my initial estimate. Other parameters are not much changing by changing Peripheral comp. Thinking this is identifiability issue (probably not), I tried a 1-comp model and the fit was worse than 2-comp and OBJFUN also much higher than 2-comp model. I tried several initial estimates of the peripheral volume but doing the same thing. I appreciate if anybody share some of their experience here. I am giving my initial estimate in this format: $THETA (0,500);POPVP Regards, Ayyappa Chaturvedula