FHRPs, like VRRP, and HSRP, don't load balance anything; that is not their purpose. The purpose of an FHRP is to give a virtual gateway, which can fail to a different physical router in the case of a primary router failure. The gateway is either statically configured on hosts, or assigned to the hosts by DHCP. Some people will also use the assigned gateway to try to spread the traffic load across multiple routers, but that can be done with or without an FHRP. That is simply an artifact of which hosts are assigned which gateway.
For example, if you have two routers and two VLANs, you can assign the hosts on VLAN 1 to use Router 1 as their gateway, and the hosts on VLAN 2 to use Router 2 as their gateway. That spreads the traffic load, but it does nothing for a failure situation.
What the FHRP will do for you is create virtual router addresses. In that case, the Virtual 1 could use Router 1 as its primary router, and Virtual 2 could use Router 2 as its primary router. That would be, under normal circumstances, no different than simply using the two routers without an FHRP. What it gives you is the ability for all the traffic to use one of the routers if the other router fails.
Some people take this even farther, and they assign multiple FHRP groups on the router interfaces. In such a case VLAN 1 could have two virtual router addresses on the single VLAN, and one of the virtual addresses could use Router 1 as its primary router, and the other virtual address could use Router 2 as it primary router. Some hosts would then be assigned to one virtual address as their gateway, and the other hosts would be assigned to the other virtual address as their gateway. This is what many people call load balancing, even though it is not really load balancing, and it has problems, especially if STP directs the frames to the root bridge, which then has to send them another direction to get to the router, resulting in sub-optimal traffic flow.
