A starting point is set on network features to define the location in the network where a trace begins. Utility network traces require that one or more starting points be defined.
Starting points are created and managed in the Trace Locations pane on the Starting Points tab. In the Trace Locations pane, you can specify terminals as starting points, differentiate between coincident features, and set multiple starting points.
Keep the following information in mind when working with starting points:
- For a starting point to be valid, the network feature it is placed on must exist in the network topology. Ensure that network features used for starting points have been validated.
- When working with spatial features on the map, a coincident point is created on the map to denote the starting location for the trace. Records from the nonspatial edge and junction object tables can be selected and added to the Trace Locations pane to serve as starting points:
- For points, devices with terminals can specify a specific terminal to set as the starting point.
- For lines, a starting point can be placed anywhere along the line feature.
- For polygons, a point is placed at the centroid to mark the starting location.
- For junction objects, features with terminals can specify a specific terminal to set as the starting point.
- For edge objects, a starting point is placed on the record to mark the starting location.
- For subnetwork-based traces, the starting points must reside in the input tier.
- When using starting points on line features, the trace results will include the entire line feature, not
just from the portion of the midspan starting point.
Note:
By default, trace results are returned as a selection and include the entire line feature. When a starting point is placed midspan on an edge feature, the Result Types option Aggregate Geometry can be specified to return partial feature results in an output multipart geometry feature class.
- If the Subnetwork Name parameter is specified on the Trace tool when performing a subnetwork trace, starting points are not used. The trace begins at the subnetwork controllers instead.
When a starting point is established, a feature class named UN_Temp_Starting_Points is created in the project's file geodatabase. Successive starting points are appended to this class. To remove all starting points from a map, use the Clear All button 
.
When an existing feature class or table is used to supply starting points for the Trace tool, the source feature class or table must include a FEATUREGLOBALID field similar to that found in the UN_Temp_Starting_Points feature class and contain a value for it to be applied to the corresponding network feature as a starting point. In scenarios in which this field is not populated, features without a corresponding value will not be used to place a starting point. This option is helpful when you want to maintain granularity over which features act as starting points.
Starting points can also be created using the Add Trace Locations tool. This tool allows selected network features or specified trace locations to be output to a feature class for use in a script or model. This can be useful to load the selection output of a trace operation as an input for a subsequent trace operation via a model.
Learn more about how to set starting points and barriers