For a Hyper-V fail-over Cluster I think if redundantly connected 12 is the recommended number of logical networks if running iSCSI last time I looked at it. (Someone feel free to correct me where I'm wrong on 3.0 best practices).
2 for Management (In VMware I just trunk this as a VLAN on the VM network, but in Hyper-V it gets weird if you don't).
2 for Storage (maybe more depending on if you need more than 2 Gbps of throughput to a host).
2 for VM's (and trunking them, possible 2 more if your DMZ requires physical separation of NICS for compliance reason).
2 for Cluster heartbeat network (their HA doesn't support storage heart-beating last time I checked, and they can't just use the management network because of the complexities of MSCS).
2 for CSV network (used because CSV's are a poor excuse of a clustered file system and needs to...