Problem

When computing a model or a rating curve in the Composite Outlet Structure manager, the following user notification is generated: "Error: Split flow not allowed for weir x-y cross section."

Solution

It is not possible to model an irregular weir that has two local low points, such as in the screenshot below.

Option 1: Flatten the high points to prevent split flow

In some cases you may only have slight undulations in the irregular section in which the split flow is only very minor. In these cases, consider modifying the irregular section to flatten any "local" high points such that there is no longer split flow. See red part in the below example case indicating a possible change to resolve this issue.

Option 2: Model multiple separate weirs in the same composite outlet structure

Another option to consider is to use a single outlet structure element and model multiple weirs within it. In the example case below you would have three weirs (or two if you were willing to flatten the high point at ~460 ft) each configured to represent the portion shown below (near-vertical "walls" at the separation points):

The introduction of two separate weirs could still skew the results slightly since the flow would pass through multiple separate weirs (two separate weir equation calculations) instead of one single weir.

Option 3: Compute rating curve externally

Another option would be to compute the rating table of the entire weir as-is using another application (if available), then input the resulting rating curve as the user defined rating table outlet option. The downside to this approach is that tailwater cannot be accounted for (free outfall assumed, since the rating curve would not be an EQT table to account for change in tailwater)

Option 4: Separate outlet structures

This approach is similar to #2 above but uses separate pond outlet nodes, pond outlet links and composite outlet structures, each configured to model the flow from one portion of the split flow. This has the same downside (plus the extra work and complication of modeling multiple outlet structures) but would enable you to direct the flow from each side of the "split" potentially to a different downstream element. Option #2 assumes that the flow splits but eventually discharges into the same common downstream element.

If in doubt, a sensitivity analysis could also be considered - try multiple options from the above list and compare the impact on the results you're interested in. If you find that they do not vary by a notable amount, then you may not need to be concerned.

See Also

Modeling a multi-stage or step weir

How do I input the elevations for an irregular weir cross section in composite outlet structures?