The nominal head and flow represent the normal operating point for the pump. In this case, it sounds like that would be having both pumps on. So, I would run the initial conditions with both pumps on and base the nominal head and flow from that (89 lps @ 52 m each, assuming same pump curve, same losses)
Behind the scenes, they are used in conjunction with the shape of the four-quadrant characteristic curves (selected by way of the Specific Speed) to construct the characteristic curve used for the pump during the transient simulation. It is expected that the nominal operating point is at or near the best efficiency point, otherwise the results may be skewed.
In a case where a pump is turned on in the initial conditions, it uses the initial conditions operating point for the nominal head and flow. When the pump is off, it is unknown, so the user must enter it.