It doesn't matter...a bonding path of one million ohms or less will prevent static build-up.
This is not necessarily true. It depends on whether the rate of charge is less than or equal to the rate of discharge. If the rate of charge is greater than the rate of discharge, you will get a static buildup. Granted, in this situation, the rate of charge is not likely to be very high. Nevertheless, the blanket statement can't always be made.
The connection between the two containers when product is being transfered in much more important than a connection to earth.
For the most part this is correct. That's because the charge buildup is actually a differential charge between the containers resulting from the transfer, and the net charge is near zero. In other words, the total charge of both containers does not change, but charge is just transferred from container to the other. The bleedoff prevents the differential from getting large.
However, if the net charge buildup for the isolated system of the two containers is non-zero, then you could have a problem. For example, if the pumping system that was used in the transfer were to introduce new charge into the sum of the other two containers, then your differential charge will stay zero, but the net charge of the whole system will be increasing.
The carpet walking example is a good analogy. Your feet would represent one container and your finger tips the other. There is no charge differential between the two, but your feet on the carpet are continually bringing new charge into the system (your body). You don't feel this new charge because there is no charge differential within the system, but when it builds high enough, you will feel the discharge when you touch the light switch.
Or, looking at it another way, the path details only matter if you expect to have a lightning strike
inside the building while transferring. And in that case you have much bigger problems to worry about.
That is not the only purpose for a grounding system, and I really wish Mike Holt would stop teaching this limited view. It is the easiest explanation to accept, but it is short sighted. There is a lot more going on than just lightning. If it was solely for lightning, then there wouldn't be the need for equipotential grids around swimming pools, and various other grounding requirements that are not lightning related (such as the topic of this thread).
Back to the OP's question:
Now for my question. Management asked me if it is better to ground in series? Ground strap from buss to drum to container being filled. Or in parallel? Ground strap from buss to drum and ground strap from buss to container being filled.
Your question is actually related to failure-mode, in that it assumes minimizing the effects of a failure of the system. If no failure exists, then it does not matter.
So the best answer to your question is to look at the most likely cause for a failure. If your ground straps are permanently connected to your ground bus bar, then that is the least likely point of failure. Even if the straps are not permanently connected to the bus, the consistency of the bus's shape still makes these connections the most stable, as compared to connecting to various irregular containers.
Therefore, your connections at each container are the weakest points of the system, and should be minimized. Regardless which methodology you use, you still have two strap-to-container connections separating the two different containers. However, in the "series connection" scenario, you actually have one additional strap-to-container connection. So you have one additional point of failure than the parallel scenario has.
So all other things being equal, the parallel scenario is better (but see my "edit" below). It also means that one of the two strap-to-container connections between the two containers is more permanent, and is not re-made every time a new container is brought into the system.
The scenario that would give you the lowest failure rate would be to have a single daisy-chained strap, where the middle connection is not dependent on two strap-to-container connections. This would be a single strap that went from the bus bar, to the first container, and then to the portable container. You still have a failure if one of those strap-to-container connections fails, but you have eliminated one of the strap-to-container connections.
Edit: In re-reading your post, my last scenario might have already been what was being asked. So yes, I agree with that part of the question.
