Looped network for gravity

I'm trying to figure out where this water all goes?
It may not be possible to model a system that doesn't have an outfall. But there are people here with more experience than I. I'm sure they will have better ideas.

Assuming you are using the default licence, which lets you use the Gradually Varied Flow solver (abbreviated to "GVF-Rational" in the software) then loops and diversions represent a special case. This is because the normal premise of System Time doesn't apply when you have a loop or diversion, which occurs when you have two conduits that carry flow out of a node. In this case, the time taken for flow to travel down each conduit will be different, and this goes against the principles of the Rational Method, where there is a single System Time at each node.

However, you can still solve this system - in two ways:

• If you stick with the GVF-Rational solver, then you need to state that one of the outgoing conduits from a node is a diversion. Do this by picking the conduit, and setting the "Is Diversion?" property to true, in the Utility properties, on the Drainage tab. You then specify a Diversion rating Curve, which is a table that state, for a given amount of flow that enters the node, a certain flow goes out through the diversion conduit (as opposed to the other - "main" conduit). You can keep these tables simple to try this out, entering two rows for example - 0, 0, and 1000, 100 - which would mean that if 1,000 units of flow enter the node, 100 units goes through the diversion. The software interpolates between the rows in the table. Note that you can only nomincate one conduit as a diversion, per node
• If you activate a CivilStorm or SewerGEMS licence (on the Tools ribbon), then you don't need to create a diversion rating table - the software will figure it out for you, as long as you use one of the solvers that computes flow over time - e.g. Explicit (SWMM Solvers). This offers more flexibility, because you can also use Control Structures, such as an orifice or a weir, to control exactly how much flow goes down each conduit. You can have multiple diversions, per node

In reality, Control Structures such as a weir will only operate when the depth in the upstream node (the HGL) reaches a certain elevation, and the operation of an orifice also depends on the upstream head. So if you want to model these types of situations, then the software needs to know how flow varies over time, which is what using the Explicit (SWMM Solvers) technique offers you. 

In either case, you will need an outfall, and I think that's in the top left of your picture. Looking at your picture, you might need a second outfall, unless the flow direction arrow on one of the conduits (the vertical one that feeds the manhole that connects to the outfall) is wrong.

Here's some links that may help:

Modeling a flow split (diversion) with the SewerCAD or StormCAD numerical solver - OpenFlows | Water Infrastructure Wiki - OpenFlows | Water Infrastructure - Bentley Communities

Using Start and Stop Control Structures for conduits and channels - OpenFlows | Water Infrastructure Wiki - OpenFlows | Water Infrastructure - Bentley Communities