Zero Lower Bound in loglinearised (by hand) stochastic model

Dear all,

I am trying to impose the Zero Lower Bound condition on the nominal interest rate in a stochastic NK model with a financial intermediation sector that I have developed. I have already log-linearised by hand as I need that for my welfare analysis of the model.

As my model is stochastic, I understand that this is not doable as in Dynare you can programme the ZLB only in a deterministic environment. I am asking if you can suggest a trick or a hint to do that in a dynamic framework which is log-linearised in advance ( I think that this might be a problem as I guess we cannot impose an inequality constraint when we take 1st and 2nd order approximations)

Thank you very much in advance,