Check a storm sewer network export for slope, cover and structure-drop problems against your own criteria with AI

Short answer

Use Copilot in Excel, Claude or ChatGPT: The AI builds check columns on your pipe and structure export (recomputed slope, flow direction, fall, structure drops, cover, depth, size changes) against criteria you typed in, and lists every row that misses one as CHECK. Storm networks get edited right up to submittal, and small problems slip through: a pipe that runs uphill after an invert edit, a short run whose total fall is too small to build, a structure with no drop, cover that went thin after a grading change. Export the pipe and structure tables from Civil 3D (or your design software) to Excel, type your agency's or client's criteria into a criteria tab, and have Copilot in Excel, Claude or ChatGPT build the check columns and a flag list by pipe and structure. Hydraulics and every design decision stay with the engineer. A person still checks: every flag in the model and profile, the criteria tab against the governing standards, and all hydraulics, utility conflicts and design decisions.

AI tie-in

Tool: Copilot in Excel, Claude or ChatGPT

The AI builds check columns on your pipe and structure export (recomputed slope, flow direction, fall, structure drops, cover, depth, size changes) against criteria you typed in, and lists every row that misses one as CHECK.

Prompt 1

Here are two tables exported from our storm network [paste]: pipes (pipe name, start structure, end structure, size, material, 2D length, start invert, end invert, reported slope, minimum cover, maximum cover) and structures (structure name, type, rim, sump, connected pipe names with their inverts). Here is my criteria tab, typed by me from our design standards [paste]. Use only these criteria and do not use any value from memory. Add: recomputed slope and its difference from the reported slope, flow direction, total fall per pipe, drop across each structure from the lowest inlet to the outlet, crown or invert match where pipe size increases, cover check, rim-to-invert depth by structure type, and size decreasing downstream. List every row that fails one of my criteria as CHECK with the value and the criterion. Never write OK or compliant.

Prompt 2

Here is my flag list [paste]. Group the flags by structure and run so I can review them in profile order from the outfall upstream, and for each group list which flags could come from the same invert edit. Do not suggest new inverts or sizes.

Human check: Every flag in the model and profile, the criteria tab against the governing standards, and all hydraulics, utility conflicts and design decisions.

The problem

Check a storm sewer network export for slope, cover and structure-drop problems against your own criteria with AI. This is a common, time-consuming task for a civil engineer.

What this workflow does

Storm networks get edited right up to submittal, and small problems slip through: a pipe that runs uphill after an invert edit, a short run whose total fall is too small to build, a structure with no drop, cover that went thin after a grading change. Export the pipe and structure tables from Civil 3D (or your design software) to Excel, type your agency's or client's criteria into a criteria tab, and have Copilot in Excel, Claude or ChatGPT build the check columns and a flag list by pipe and structure. Hydraulics and every design decision stay with the engineer.

How it works: AI drafts or analyzes. A person operates the tools, checks the work, and takes action.

Your checkpoint: The sheet only checks geometry against criteria you entered. It does not check hydraulics, utility conflicts or constructability in the field, and the engineer of record reviews every flag and every design decision. Confirm the export matches the model before relying on it.

Before you start

Access to the named tool; confirm license, version and cost with the vendor.
A small, non-confidential sample to try the workflow on first.
A qualified person to review the output before it is used.

Tools and access

Follow the steps

Export the network to Excel: a pipe table (pipe name, start and end structure, size, material, 2D length, start and end invert, slope, minimum and maximum cover) and a structure table (structure name, type, rim, sump, and the invert of every connected pipe). A pipe table and structure table in the drawing exported to CSV, or the Prospector part lists copied out, both work.
Make a criteria tab and fill it in yourself from the design standards that govern the job: minimum and maximum slope by pipe size, minimum cover by surface type, minimum drop across a structure, whether crowns or inverts are matched when pipe size goes up, the maximum depth for each structure type, and the smallest total fall per run your contractors can reliably hold. Tell the AI not to supply any of these values from memory.
Have Copilot in Excel (or a formula set written by Claude or ChatGPT) add check columns: slope recomputed from inverts and length next to the reported slope, flow direction (downstream invert lower than upstream), total fall per pipe, drop across each structure from the lowest inlet to the outlet, crown or invert match where size changes, cover against the criteria for that location, rim-to-invert depth against the structure type, and pipe size getting smaller downstream.
Have it produce one flag list: pipe or structure ID, the check, the value, the criterion it was compared to, and CHECK. It never writes OK, passes or compliant.
Review every flag in the model and the profile view, fix in Civil 3D, re-export and re-run. Keep the criteria tab and each run's flag list with the submittal it went with.
Capacity, velocity and hydraulic grade line come from your hydraulic analysis, not from this sheet. Utility crossings and separation still need their own check against the utility plans.

Try a small example

Replace the bracketed fields with your own project details. No tested sample run is published for this workflow yet.

Sample input

Fill in the bracketed [fields] in the text below with your project data.

Starter prompt

Here are two tables exported from our storm network [paste]: pipes (pipe name, start structure, end structure, size, material, 2D length, start invert, end invert, reported slope, minimum cover, maximum cover) and structures (structure name, type, rim, sump, connected pipe names with their inverts). Here is my criteria tab, typed by me from our design standards [paste]. Use only these criteria and do not use any value from memory. Write Excel formulas or a step-by-step Copilot plan that adds: recomputed slope and the difference from the reported slope, flow direction, total fall per pipe, drop across each structure from the lowest inlet to the outlet, crown or invert match where pipe size increases, cover check, rim-to-invert depth check by structure type, and size decreasing downstream. Then list every row that fails one of my criteria as CHECK with the value and the criterion. Never write OK or compliant.

Expected output

Not published yet. Treat any output as a draft and check it against the source documents.

Check the result

Every value, citation or quantity traces back to a source document.
Missing information is marked as missing, not filled in.

Limits and recovery

The sheet only checks geometry against criteria you entered. It does not check hydraulics, utility conflicts or constructability in the field, and the engineer of record reviews every flag and every design decision. Confirm the export matches the model before relying on it.
Check your organization’s data policy before sharing project files with third-party tools.

Stop or switch when: The source cannot be verified, permissions are missing, or the result contains unsupported facts.

Frequently asked questions

How does AI help with this?

Copilot in Excel, Claude or ChatGPT: The AI builds check columns on your pipe and structure export (recomputed slope, flow direction, fall, structure drops, cover, depth, size changes) against criteria you typed in, and lists every row that misses one as CHECK. A person still checks: Every flag in the model and profile, the criteria tab against the governing standards, and all hydraulics, utility conflicts and design decisions.

How do I check a storm sewer network export for slope, cover and structure-drop problems against your own criteria with AI?

Export the network to Excel: a pipe table (pipe name, start and end structure, size, material, 2D length, start and end invert, slope, minimum and maximum cover) and a structure table (structure name, type, rim, sump, and the invert of every connected pipe). A pipe table and structure table in the drawing exported to CSV, or the Prospector part lists copied out, both work. Make a criteria tab and fill it in yourself from the design standards that govern the job: minimum and maximum slope by pipe size, minimum cover by surface type, minimum drop across a structure, whether crowns or inverts are matched when pipe size goes up, the maximum depth for each structure type, and the smallest total fall per run your contractors can reliably hold. Tell the AI not to supply any of these values from memory. Have Copilot in Excel (or a formula set written by Claude or ChatGPT) add check columns: slope recomputed from inverts and length next to the reported slope, flow direction (downstream invert lower than upstream), total fall per pipe, drop across each structure from the lowest inlet to the outlet, crown or invert match where size changes, cover against the criteria for that location, rim-to-invert depth against the structure type, and pipe size getting smaller downstream. The full step list is above.

Does this use an AI agent or regular software?

It is labeled AI assists: AI drafts or analyzes. A person operates the tools, checks the work, and takes action.

What tools do I need?

Autodesk Civil 3D, Microsoft Excel, Microsoft Copilot, Claude, ChatGPT. Confirm license, version and cost with the vendor, and try it on a small non-confidential sample first.

Has this workflow been tested?

Workflow guide · not independently tested. Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written.

Evidence and maintenance

Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written.