I cannot for the life of me find a resource which gives the general state space form for an ARIMA(p,d,q)(P,D,Q) model. I am reading Time Series Analysis by State Space Methods, by Durbin & Koopman, which contains the general form for ARIMA(p,d,q) along with some examples.
However, I'm trying to use the statespacer R package and I am struggling to grasp how a (0 1 1)(0 1 1)_12 model would have 27 state components (just going by the size of $T$ when I run an example in one of the vignettes). I have tried to extrapolate based on the examples, and even looked at a $(0 1 1)(0 1 1)_4$ example in a different book by Pelagatti, but still can't grasp what the general form should be...
I also checked some resources on Google but was unable to find what I'm looking for. Any help pointing me in the right direction would be appreciated!