by Provision
Healthcare construction specs regularly exceed 2,400 pages. A plumbing estimator has two days — maybe three — to price the job. That math doesn't work. Not manually.
The result isn't just slow turnaround. It's missed requirements. Scope gaps that show up as change orders six months into the project. Or worse, costs absorbed quietly under "readily inferable" contract language.
This article breaks down where plumbing scope gaps actually come from, what a thorough Division 22 review requires, and how AI is changing what's possible for estimators on spec-heavy projects.
Plumbing scope doesn't live in one place. Division 22 is the starting point, but plumbing requirements scatter across other spec sections, general conditions, supplementary conditions, civil drawings, mechanical drawings, and addenda.
A fire-rated floor penetration might appear in Division 07. A generator domestic water connection might be buried in Division 26. The trench — who digs it, who backfills it, who owns the compaction — might not appear in Division 22 at all.
This is the core problem. Plumbing scope is cross-referenced, and estimators reviewing only one section of a 2,400-page spec book are working with incomplete information.
The trench is one of the most consistently disputed scope items in MEP work. Civil and plumbing drawings often assume the other trade is responsible. Specs rarely make it explicit. By the time it surfaces in the field, both subs have a reasonable argument for why it's not their work.
The GC absorbs the cost — either by directing one sub to do the work or by eating the delay while the dispute gets resolved. Either way, it's a scope gap that a thorough pre-bid review would have caught.
Generator domestic water supply and field conditioning is another recurring gap on healthcare projects. The gap appears because mechanical and electrical drawings often cross-reference each other without explicitly assigning the work. Estimators miss it. Then it shows up as a claim. According to interviews conducted for The Scope Gap Playbook, "millions" in disputed generator field-conditioning costs have recurred across multiple healthcare and institutional projects at the same firm.
That's not a one-time oversight. That's a systemic review problem.
Most estimators know the core Division 22 checklist. Sanitary drainage. Domestic hot and cold water. Natural gas. Medical gas rough-in on healthcare projects. Storm drainage. Plumbing fixtures and trim.
What gets missed is the layer underneath — the items that require cross-referencing multiple documents to even identify.
None of these items are exotic. All of them get missed on jobs reviewed manually under bid-day time pressure.
The $300K lead-lined glass omission that cost one GC on a hospital imaging suite is the kind of example that gets attention at post-mortems. But plumbing gaps are usually smaller and more frequent — which makes them harder to track but just as damaging to margin.
According to the Scope Gap Playbook — built from interviews with 200+ general contractors — "If you miss anything, they'll bill it" is how a Pre-Construction Lead at a top-ENR Canadian GC described today's subcontractor environment. The era of the gentleman's agreement is over.
An Estimating Manager at a Canadian ICI GC put it plainly: "We have less subs who just kind of [have] a gentleman's agreement… they've become more quick to clarify that we're not including that one piece of scope."
Subcontractors are getting more precise about what they include. GCs need to match that precision in the scope they issue — or they own the gap.
The FMI Construction Disconnected report puts annual U.S. rework costs from miscommunication and bad project data at $31 billion. Twenty-six percent of that rework traces back to communication breakdowns. Twenty-two percent to bad project data — specs that were misread, missed, or never reviewed.
The Arcadis 2025 Global Construction Disputes Report puts the average U.S. construction dispute value at $60.1 million. "Errors and omissions in contract documents" has been the number-one dispute cause for six of the last nine years.
Plumbing scope gaps don't always generate $60M disputes. But they feed the same root cause: requirements that existed in the documents and weren't surfaced before execution started.
The problem isn't that estimators are careless. The problem is volume and time.
A 2,400-page healthcare spec book contains plumbing-relevant requirements in a dozen or more sections. An estimator reviewing it manually — even a good one — is making judgment calls about which sections to read in depth and which to skim. Under bid-day pressure, that means some requirements don't get reviewed at all.
The anti-pattern the Scope Gap Playbook calls "just send the bid docs" captures the worst version of this: scope packages that reference the full spec set without surfacing any specific requirements. Subcontractors receive a 2,400-page PDF and bid what they know. The GC assumes coverage. The gap lives quietly until it surfaces as a change order.
Another common failure is what the Playbook calls "the five-minutes-before-bid review" — a rushed check at the end of the process that can't realistically catch cross-referenced items buried in supplementary conditions or late addenda.
With $48B in healthcare construction started in Q1 2026, the spec volume problem isn't going away. Healthcare specs are longer, more detailed, and more heavily cross-referenced than typical commercial work. Medical gas systems, infection control requirements, seismic bracing, commissioning — every one of these adds scope that exists somewhere in the document set.
Healthcare owners also have less tolerance for scope disputes and change orders mid-project. The operational stakes are higher. That puts more pressure on the GC to get the scope right before execution starts.
Purpose-built construction AI doesn't replace the estimator's judgment. It handles the part of the job that doesn't require judgment: reading every section of a 2,400-page spec book and surfacing every plumbing-relevant requirement before the estimator starts pricing.
Provision's Scope Agent ingests the full project set — drawings, specs, contracts, addenda — and generates a complete scope-of-work package in under 60 minutes. That's 30–40 hours of manual review compressed into one hour, with requirements pulled from every relevant section of the document set, not just Division 22.
This matters for plumbing specifically because of the cross-referencing problem. Scope Agent reads Division 07, Division 23, Division 26, and the supplementary conditions alongside Division 22. Requirements don't get missed because no one had time to read that section.
Even a thorough scope extraction can miss contractual risk buried in general conditions. That's where Risk Review comes in.
Risk Review runs against the full contract and spec set using pre-built risk checklists built for GC pre-construction. On Division 22 and MEP work, that means flagging items like:
Risk Review carries 99.5% accuracy on pre-built checklists. For custom project-specific checklists, accuracy holds above 97%. That's a material difference from running a healthcare spec through a generic AI tool and asking it to "find the risks."
Provision has reviewed over $100 billion in project value and processed more than 66,000 documents — including complex healthcare and institutional project sets. The EllisDon case study shows what that looks like in practice on a high-complexity project: $1.8M in identified scope gaps before execution started.
Scope packages and risk checklists cover the structured review. But estimators also have one-off questions that come up mid-bid. Who owns the concrete pad under the domestic water booster pump? What does Division 22 say about the warranty on plumbing fixtures?
Provision's Chat Agent answers those questions directly, with citations to the specific document and section. No searching. No scrolling through a 2,400-page PDF. An answer in under 20 seconds, sourced from the actual project documents.
Provision has answered over 50,000 queries across real project document sets. The accuracy rate across those queries is 95% — verified against the source documents.
The Scope Gap Playbook identifies eight habits that separate high-margin GCs from the rest. Several apply directly to how plumbing scope gets written and issued.
Habit 1: Drawings-first, not boilerplate-first. The plumbing scope should start from what the drawings actually show — not a template pulled from the last hospital project. Two healthcare projects rarely have the same medical gas layout, the same fixture schedule, or the same commissioning requirements.
Habit 2: Specific document references, not generic incorporation. "As per plans and specs" is the most dangerous phrase in a plumbing scope package. It transfers no information. It assigns no responsibility. It just creates a larger dispute later. Specific references — "see Drawing P-201, Section 220500 Paragraph 2.3" — are harder to argue with in the field.
Habit 3: Templates as a floor, not a ceiling. A good scope template captures the standard items. It doesn't capture project-specific requirements from a 2,400-page healthcare spec. Templates need to be supplemented with a thorough review of the actual document set. That's where AI earns its value.
Habit 4: The pre-issue scope review checkpoint. Before the plumbing scope package goes to subs, someone should check it against the spec book — specifically looking for requirements that were missed. On a 2,400-page healthcare project, that check takes hours manually. With Scope Agent, it takes less than one.
For a deeper dive on how these habits apply across all trades, see the Trade-Specific Scope Gaps chapter of the Scope Gap Playbook — and the scope of work template for a starting point on structure.
A chief estimator at a GC working a $120M hospital project gets the spec book — 2,600 pages — with 18 days to bid. The plumbing sub needs a scope package in five days to price accurately.
Manually, building that scope package takes 35 hours. Requirements scattered across Division 22, Division 07, Division 23, Division 26, and supplementary conditions. Three estimators, five days, still risks missing the generator field conditioning requirements buried in a mechanical addendum issued on day 14.
With Provision, the full document set is uploaded once. Scope Agent generates a complete plumbing scope package — cross-referenced across every division — in under 60 minutes. Risk Review flags the commissioning participation language and the "readily inferable" clause in the supplementary conditions that the manual review would have missed.
The estimator spends the five days on pricing, not reading. The scope package that goes to subs is complete. The risk is surfaced before bid day, not six months into the project.
That's the difference between a 30-hour manual review and a one-hour AI-assisted review for general contractors doing MEP-heavy commercial work.
To see how this works on a real healthcare or institutional project set, request a demo.
A complete review covers Division 22 plumbing specs, but also cross-referenced requirements in Division 07 (penetration fire-stopping), Division 23 (pipe insulation), Division 26 (motor starters and disconnects), supplementary conditions, and all addenda. Healthcare projects routinely include medical gas, seismic bracing, and commissioning requirements that scatter across multiple spec sections.
Plumbing requirements are spread across multiple spec divisions, drawings, and addenda. Manual review under bid-day time pressure leads to incomplete coverage. Items like the trench ownership, generator field conditioning, and fire-stopping responsibilities are frequently missed because they require cross-referencing documents no single estimator has time to read in full.
Scope Agent ingests the full project set — drawings, specs, contracts, and addenda — and extracts all plumbing-relevant requirements, not just those in Division 22. It generates a complete scope package in under 60 minutes, replacing 30–40 hours of manual review. Requirements from supplementary conditions and cross-referenced sections are included automatically.
Scope Agent generates the scope-of-work package — what plumbing work is required on the project. Risk Review analyzes the contract and specs for risk exposure — ambiguous trench language, "readily inferable" clauses, commissioning obligations, and liquidated damages tied to MEP milestones. Both tools are used together on complex projects.
Provision's tools are built for exactly this. Provision has processed over 66,000 construction documents and reviewed more than $100 billion in project value. Healthcare and institutional project sets — including multi-volume spec books with extensive addenda — are standard inputs. The Chat Agent can answer specific questions from that document set in under 20 seconds, with cited references.
The most consistently missed items are: trench ownership, generator domestic water and field conditioning, motor starters for equipment connections, seismic bracing, fire-stopping at penetrations, concrete equipment pads, temporary construction-phase plumbing in Division 01, and commissioning participation requirements. All are identifiable in the documents — they just require reading sections estimators don't always reach under time pressure.
"Readily inferable" clauses in general conditions can assign plumbing scope to the GC — or push it to subs — without explicit specification. A $300K lead-lined glass omission on a hospital imaging suite was absorbed under exactly this language. Identifying and flagging these clauses before bid day is a key function of Provision's Risk Review tool.
Provision was built specifically for GC pre-construction. Not a general-purpose model retrofitted for construction. See the difference in a 30-minute demo.
Book a demoMore Articles