Between contract signature and the first crate arriving on site sits the step most kitchen projects underestimate: the shop drawing and submittal cycle. It looks like paperwork. In practice it is where airflow assumptions get locked in, where responsibilities get assigned on paper, and where a programme either holds or slips week by week. Once you stamp a drawing approved, you have accepted every number on it: the airflow, the static pressure budget, the electrical loads, and the weights your structure must carry. This guide is written for the buyer side: restaurant groups, dealers, project managers and contractors who receive a supplier’s submittal package and must decide, line by line, whether it is safe to approve.
Why the Approval Stage Decides the Project
In most supply contracts, approval of drawings is a contractual trigger: fabrication time and delivery lead time start counting from the date the buyer returns approved drawings, not from the date of the purchase order. That single clause makes the review loop a programme instrument. If the package sits on a desk for three weeks, or if it comes back for a second and third round of comments, those weeks reappear at the end of the job as a compressed installation window or as grounds for a delay claim. Treating the submittal cycle with the same seriousness as the price negotiation is therefore not bureaucracy; it is schedule control.
Approval also allocates risk in the other direction. When the buyer stamps a drawing approved, the supplier gains a documented record that the buyer accepted the configuration shown. If the hood later fails to fit the survey, or the fan proves undersized, the approved drawing becomes the supplier’s defence. The discipline that protects the buyer is simple: review before approving, comment in writing, and never approve verbally or by implication. What you do not write down, you have not reviewed.
One boundary should be stated in your own review procedure, because suppliers rarely state it for you: the buyer’s approval is a coordination approval, not a design warranty. You are confirming that the layout fits the building, that the airflow basis matches the agreed duty, and that the interfaces line up. The technical adequacy of the equipment inside the box remains the specialist supplier’s responsibility under the contract. Writing this distinction into your transmittal letters prevents a later argument about who owns a design error.
What a Complete Submittal Package Contains
A kitchen exhaust submittal is not one drawing. It is a bundle of documents, each proving a different thing, and a missing document is itself a review finding. Before any detailed checking, verify that the bundle is complete:
| Document | What it proves |
|---|---|
| General arrangement and sections | Hood dimensions, mounting height, overhang beyond the cookline, clearances for opening doors and access panels |
| Duct routing and sizing | Duct diameters, velocities, route through the building, penetration and fire-stopping locations, access panel positions |
| Fan selection with curve | The operating point plotted on the fan curve, total pressure, motor power, sound data |
| Purification stage details | ESP cell construction and count, collector plate material, high-voltage power supply, activated carbon quantity for the odour stage |
| Electrical schematics | Single-line diagram, breaker sizes, voltage and frequency for the market, interlock wiring between hood, fan and purification stages |
| Structural loading sheet | Unit weights, hanging loads, roof or wall fixing reactions for the structural engineer to verify |
| Coordination interfaces | Fire suppression layout over the cookline, gas interlock, make-up air route, condensate drainage |
| Draft O&M and commissioning plan | Maintenance access requirements, planned test procedure, the acceptance numbers the site will be measured against |
If any of these arrive as a promise rather than a document, log it. Packages that are 90 percent complete consume the same review window as complete ones, and the missing ten percent is usually the part that carries the interfaces.
The Numbers to Recompute, Not Just Read
The most common review failure is reading the airflow figure and nodding. The figures were derived by the supplier from your duty description; recomputing them from first principles takes ten minutes and catches most sizing errors before steel is cut.
Airflow cross-check. Take the hood length from the drawing and apply the two standard algorithms: roughly 2,000 m3/h per metre of hood for enclosed cooklines, 2,500 m3/h per metre for open display cooking, or about 4,000 m3/h per gas burner. The two results should agree within about 30 percent. A worked example: a 3 metre hood over an open display grill line gives 3 x 2,500 = 7,500 m3/h by length; if the line carries two burners, the per-burner check gives 2 x 4,000 = 8,000 m3/h. The two agree within 7 percent, so the stated 7,500 to 8,000 m3/h duty is defensible. If the drawing states 5,500 m3/h, you have found a problem worth a formal comment before approval, not a discussion after installation.
Face velocity. Divide the stated airflow by the hood’s open face area. Capture normally requires a face velocity in the range of 0.4 to 0.5 m/s at the hood edge. A large hood with a small fan passes the total airflow check and still fails this one; spillage at the ends is the symptom.
Static pressure budget. Ask for, or build, the resistance ledger that the fan must overcome:
| Component | Typical contribution |
|---|---|
| Hood filters and grease baffle bank | 50 to 120 Pa, rising with loading between cleaning cycles |
| Purification stage (ESP cell plus carbon, if in-line) | 80 to 200 Pa depending on stage count and media face velocity |
| Duct friction over the full run | Roughly 0.8 to 1.2 Pa per metre at normal grease-duct velocities, plus fittings |
| Silencer and discharge louvre | 30 to 80 Pa combined |
| Design margin | 10 to 15 percent on the sum, for filter loading growth |
The fan’s duty point on its curve must sit at the total airflow against this total pressure. A fan named by model number without a plotted curve is a red flag on its own: you cannot verify power draw, you cannot verify the operating point, and you cannot later diagnose whether poor capture is a fan problem or a resistance problem. Motor power at the duty point is also the number your running-cost estimate needs, so record it.
Purification stage substance. Where the package shows a purification stage, insist on construction detail rather than a labelled box: the number of ESP cells against the airflow, the cell face velocity, the collector plate material, the type of high-voltage power supply, and for the carbon odour stage the kilograms of activated carbon and the stated contact time. These are the exact items that reappear in the acceptance test and in the consumable budget, so a submittal that hides them is a submittal deferring an argument to the worst possible moment, which is commissioning.
Electrical and structural reality. Check that the schematic matches the market: 230 V / 50 Hz for most of Europe and Asia, 120 or 240 V / 60 Hz for the Americas, and confirm breaker sizes and cable routes against the site’s distribution board. Check that weights are stated per unit and per fixing point; the structural engineer cannot review a hood that carries no weight table, and discovering a roof upgrade requirement after approval is a cost nobody budgeted.
Approval Codes and What They Actually Commit You To
Every review must end in a code, and the codes are not interchangeable. Ambiguous stamps are how errors travel into fabrication wearing your signature:
| Code | Meaning and buyer action |
|---|---|
| Approved / Approved for construction | Fabrication may proceed on this revision. Use only when every comment is resolved and the revision number is stated on the stamp. |
| Approved as noted | Proceed, but the noted comments become part of the contract documents. Each note must be specific and numbered; a stamp with vague notes transfers the argument to you. |
| Revise and resubmit | Work may not proceed. List every required change with a drawing reference; a single-word rejection invites a resubmission that fixes only one of them. |
| For coordination only | Not a fabrication release. If a supplier later fabricates from a coordination sheet, that is their risk to carry, and it should be said in writing. |
State the code, the date, the reviewer and the revision number on every returned sheet. A review without a revision reference is untraceable, and untraceable reviews are what both sides argue about in month eleven.
Eight Traps Hidden in Submittal Packages
Across projects, the same failures recur. Each is cheap to catch in review and expensive to catch on site:
1. Catalog cuts instead of project drawings. A brochure page is not a shop drawing. It shows a unit that exists; it does not show the unit that fits your kitchen, your ceiling height or your duct route. Require dimensioned project-specific sheets.
2. Coordination sheets used for fabrication. Sheets marked “not for construction” that quietly become the fabrication basis. Watch the revision stamps when crates arrive.
3. Fan without a curve. Covered above, and worth repeating: a fan selection you cannot plot is a fan selection you cannot accept.
4. Generic electrical diagrams. A single-line diagram with no breaker sizes, no terminal references and no interlock wiring will be redrawn by the site electrician on installation day, which is where improvisation belongs to nobody.
5. Missing weights. No structural sheet means no structural review. Roof-mounted fans and suspended hoods concentrate load on a few fixings; the engineer needs reactions, not adjectives.
6. The purification stage as a black box. No cell count, no carbon mass, no power supply detail. This defers the acceptance basis to the supplier’s own commissioning report, which is circular.
7. Ambiguous dimensions. Notes such as “refer to architect” or “confirm on site” inside a fabrication drawing push design decisions into the installation window. Every dimension on a shop drawing should be a number or a documented allowance.
8. Revision chaos. Sheets without revision clouds, without revision numbers, or mixed from two revisions in one package. If you cannot tell which sheet is current, neither can the workshop.
Turning the Approval Loop into a Programme Discipline
The review process itself needs a shape, or it will consume the schedule it was meant to protect. Four habits make the loop short and traceable:
A submittal register. One table, maintained by one person, listing every document with its revision, submission date, reviewer, due date, returned code and return date. Ten working days is a workable review turnaround for a package of this size; anything longer should be logged as a programme event, because under most contracts it becomes one.
One reviewer, consolidated comments. Comments split across five people arrive as five letters in five weeks. Route the package through one technical reviewer, collect internal comments, and return one letter. Partial responses trigger partial resubmissions, and partial resubmissions multiply cycles.
A freeze point. After the code reads approved, the configuration is frozen. Later changes, whoever requests them, go through the variation procedure with their programme and cost consequences stated. An approval that can be reopened by conversation is not an approval.
Evidence discipline. File every transmittal, every comment letter and every stamped sheet under the revision number. When a dispute arrives, the file either answers it in an afternoon or it does not exist.
From Approved Drawings to Handover Records
The submittal discipline does not end at approval; it converts at handover. The package you approved becomes the as-built record once marked up with what was actually installed, and that record is what the maintenance contractor, the insurer and any future retrofit will read. Three additions close the loop. First, require record drawings as a deliverable, not a favour, with duct routes, fixing points and terminal locations as installed. Second, have commissioning readouts written onto the record set: airflow at the duty point, fan motor current, ESP cell voltage and current, face velocity, and noise at the agreed reference positions. These become the baseline that the defects period is measured against. Third, require the draft O&M manual to be completed as a final document, with the consumable schedule and the spare parts list consistent with what is actually installed. A record set with numbers is an asset; a record set of approved drawings without readouts is only paper.
How Architecture Changes the Submittal Stack
The volume of the approval loop is not fixed; it follows from the architecture chosen at selection stage, and this is where the earlier decision work pays off. A ductless all-in-one configuration removes the documents that generate most review cycles: there is no duct routing sheet, no building penetration detail, no fan curve against a long duct run, and no make-up air balancing package, because the unit captures, purifies and recirculates within one enclosure. The submittal shrinks to a general arrangement, an electrical schematic, a weight sheet and the purification stage detail, and the loop that took six weeks takes two. For kitchens where the building cannot accept new penetrations, the ductless range hood range is built around exactly that contraction of interfaces, pairing an electrostatic stage with an activated carbon odour stage inside the hood.
Within the ductless route, the menu still picks the unit. Light-fume lines such as coffee, griddles and steaming fit a compact all-in-one hood; the N-series serves that tier. Heavy-fume lines with fryers and char-grills need the deeper cell bank and larger carbon stage of the ND-series, and where odour limits are tight, a separate Odor Removal Tower in the supply path adds a second treatment stage without adding a duct run. The submittal review for these units concentrates on the same items a ducted package would: cell construction, carbon quantity, electrical detail and weights, just with fewer documents around them.
Where the building does support a full duct run and the duty is high-volume, a ducted line remains the engineering answer, and the submittal review carries the full weight described above: the ducted ESP purifier range covers that end, with the fan curve, the static pressure budget and the ESP cell construction as the three core review items. For cooking that moves, including market pitches, front-of-house live stations and pop-up counters, pre-engineered mobile units reduce the approval question to one of configuration: the ventless catering cart line ships as an engineered package whose review reduces to dimensions, power and certification documents.
Across all three routes, the same rule holds: the more of the system that arrives as pre-engineered factory assembly, the smaller the submittal stack and the shorter the loop, but the fewer documents there are, the more each one matters. Review them all.
A Ten-Point Approval Checklist
Print this and keep it with the submittal register:
1. Bundle complete: all eight document classes present, or gaps logged with dates.
2. Drawings are project-specific, not catalog cuts, and carry a revision number.
3. Airflow recomputed by both algorithms, agreeing within 30 percent.
4. Face velocity between 0.4 and 0.5 m/s at the stated face area.
5. Fan duty point plotted on a curve, with total pressure and motor power stated.
6. Static pressure ledger sums to the fan’s total pressure, with 10 to 15 percent margin.
7. Purification stage specified in construction terms: ESP cell count, collector plates, power supply, carbon kilograms and contact time.
8. Electrical schematic matches market voltage and frequency, with breaker sizes and interlocks.
9. Weights and fixing reactions stated for structural review.
10. Approval stamped with code, date, reviewer and revision, and the configuration frozen thereafter.
Approving kitchen exhaust drawings is not a clerical step on the way to delivery; it is the moment the project’s numbers become commitments. Review the bundle for completeness, recompute the airflow and pressure figures, read every approval code as the contract it is, and convert the approved set into a record with commissioning numbers at handover. Buyers who run that discipline get installations that match the drawings, programmes that survive the approval loop, and a paper trail that answers disputes instead of starting them.
