Guide · Piping take-off
How to get a piping MTO out of Navisworks
Short answer: Navisworks can do it. Quantification ships with Simulate as well as Manage, and it exports to Excel. What it does not do is calculate anything from geometry — it reads what the authoring application already wrote into the model. So the real question is never “can the tool count?” but “did the model publish enough for anything to count?” and “how many estimator-days does the set-up take?”
What Navisworks actually gives you
Autodesk’s own Quantification workflow is eleven steps: open the source file, open the workbook, set up the project, create or select take-off items, hide what you do not want, measure what the catalogue misses, edit items, edit formulas and parameters, refresh when the data changes, validate, and export to XLSX. Underneath those steps sit two libraries you build yourself — an Item Catalogue that defines what you are measuring, and Property Mapping that connects a model property to a take-off field.
Search Sets make this far less manual than it sounds. A rule-based set can be dragged onto a catalogue item and take off every matching object in one move. If your firm builds a good catalogue and set library once, most of that cost disappears on the tenth project. This is worth saying plainly: a mature Navisworks template is genuinely productive, and any comparison that pretends otherwise is not being honest with you.
Where the work actually is
The bottleneck is not the licence. Navisworks Simulate is around £860 a year excluding VAT in the UK — roughly £72 a month for a business that reclaims VAT. That is not what makes a take-off expensive. What makes it expensive is the estimator-days: the person who builds the catalogue, maps the properties, distributes items to the right WBS, checks the result, and does it again when a revision lands two days before the bid.
Autodesk University’s own class description calls take-off one of the most time-consuming parts of estimating. That is the line item to measure — not the seat.
Three things that quietly go wrong
1. Zero-length pipes
Quantification’s ModelLength, ModelWidth and ModelVolumeare, in Autodesk’s own words, properties that came into the model from the original design application. Navisworks does not derive them from the solid. When Plant 3D pipes come through with length reading zero, this is why: the property has to be mapped by hand before the number appears.
2. Duplicates in a federation
Nothing in Quantification de-duplicates identical objects. Duplicate detection is a Clash Detective feature, and Clash Detective is Manage-only — so a Simulate user has no tool for it at all. Meanwhile Autodesk documents that appended NWCs can multiply in an NWF on open or refresh. The failure mode is nasty because it is invisible: totals inflate and the 3D view looks perfectly normal.
Measured case. In one 210 MB federation we processed, 82,200 objects were exact duplicates. Counted naively they would have added roughly 7,104 m of pipe that does not exist. The model was not broken — federations simply accumulate copies, and no one had a tool pointed at it.
3. Disciplines bleeding into pipe totals
Duct elbows, insulation sleeves and cable containment sit next to pipework in the tree and look similar to a rule that matches on shape or name. Without an explicit discipline gate they end up inside pipe quantities, and the error survives all the way to the price.
What the model has to publish
No tool — Navisworks, Metriq, or anything else — can count what the model does not describe. Before you judge any take-off software, check the model for:
- A class or category on each component (Pipe, Elbow, Flange, Valve, Tee, Reducer).
- A size that resolves to a nominal bore — ideally an OD/DN pair, not a bare number. A bare “40” can be a schedule, a pressure class, or a millimetre bore.
- A length on pipe segments, published by the authoring tool.
- A line or system name, if you want the take-off broken down by line rather than by one flat total.
If those are missing, the honest output is a shorter list plus a written statement of what could not be measured. An output that fills the gap with estimates is worse than no output, because it prices work nobody can trace.
A sane output contract
One complaint shows up again and again in forums: the Quantification export gives every dimension for everything, when what an estimator wants is length for pipe and count for everything else. Whatever tool you use, hold the output to a contract:
| Item | Unit | Why |
|---|---|---|
| Pipe | metres | Bought by length; cut lists come later |
| Elbows, tees, reducers, caps | each | Bought as discrete parts |
| Flanges | each | Type matters (weld neck / slip-on / blind) |
| Valves | each | Often priced separately from bulk |
| Reducing tee / reducer | each, two sizes | A single size makes the row unpurchasable |
The two-size rule is the one most often broken. A reducing tee with a blank second size cannot be ordered, and the mistake is easy to miss because the row looks complete. Whether a fitting is reducing is decided by measurement, not by the family name — catalogue names like “Reduced” appear on equal tees too.
How to check any take-off before you trust it
- Count the objects twice. Take an independent count straight from the model tree and compare object counts, not just totals. If the object counts match and the metres do not, the problem is a property, not a missing part.
- Look for the things that are not there. Ask what the tool could not classify. A take-off with no “unresolved” list is usually hiding one.
- Trace three rows back to the model. Pick a large quantity and follow it to the objects it came from. If you cannot, you cannot defend the number in a meeting.
- Check the diameter bands. Band-by-band agreement is far stronger evidence than a matching grand total, which can hide two errors cancelling out.
Metriq MTO does this pass automatically: you upload the NWD, choose which source models are in scope, and get a line-based piping MTO in your client’s own Excel schema — with duplicates counted once and anything unmeasurable listed rather than invented.