You would have spikes in transaction volumes at the Primary whereby the outstanding redo / archivelog to propagate to the Standby exceeds the network bandwidth. This results in the Standby lagging as the network is inadequate (OR the standby disks are much slower than the primary so the standby is unable to handle the spike in transaction volume). In most cases, it is that the network cannot handled the spikes in transaction volume.
Hemant K Chitale
Does your standby databases have Standby Redo Logs? Or just regular Online Redo Logs? You can verify by query V$STANDBY_LOG in the standby database.
I bet you don't have SRLs which is why you see this activity. Get them. They are your friend. If you confirm that you don't have SRLs, I can expand more on the topic.
In the shop I work at the network and the SAN both are used for other things. Since we don't control the load on either they could cause lag.
If you have not seen this its worth a look: