It is hard to tell. The problem often is that there is still a mistake somewhere in the model equations and steady state computations then fail, because the wrong equations do not have a steady state. Given the size of your model, it will be very hard to debug. I can only recommend trying to compute the steady state analytically.