The stages of a lighting design commission, and what leaves your desk at each one

Most lighting schemes are not lost in design. They are lost in the gap between design and delivery — a substitution nobody flagged, a ceiling void that shrank, a commissioning visit that was never in anyone’s fee.

Knowing what leaves your desk at each stage, and what must be locked before it does, is most of the job.

Concept

You produce: narrative, references, a lighting strategy diagram, indicative sections showing the key moves, mood imagery, an order-of-cost band.

You must lock: the design intent, and the decisions that only the architect can give you — ceiling voids, cove depths, recess allowances, the presence or absence of a slot.

This is the only stage where you have real influence over the building’s geometry, and geometry sets the ceiling on what any lighting scheme can achieve. A 40 mm cove detail agreed at concept is worth more than any fitting selection made later. Fight for the details now; there is no second chance at this.

Do not select products at concept. Describe effects and the performance the architecture must support.

Schematic design

You produce: a fitting-type schedule with performance specifications, indicative layouts, first calculations for the critical spaces, a control strategy diagram, a budget estimate.

You must lock: ceiling types and construction, structural coordination for anything recessed or suspended, the control philosophy, and the standards you are designing to.

Two things belong here that are almost always late. Write down which standards apply — EN, IES, local code, WELL, LEED, planning conditions — because they determine the numbers and because discovering a mandatory requirement at tender is expensive. And run one calculation early for the hardest space in the project, to find out whether the concept is physically achievable before it is presented as fact.

Detailed design

You produce: full layouts, complete schedules, calculations for every regulated space, mounting and installation details, the scene schedule, the specification, an emergency lighting strategy.

You must lock: final surface reflectances, final furniture and task locations, coordination with mechanical services in the ceiling void, and the maintenance strategy.

Re-run the calculations against the final finishes. Reflectance assumptions made at schematic are usually optimistic, and the interior designer has almost certainly changed something. This single step catches more problems than any other in the process.

Write the scene schedule here — room, scene name, groups, levels, CCT, fade time, trigger. It becomes the commissioning brief, and without it commissioning becomes improvisation.

Tender

You produce: the tender specification, the approved-equivalent criteria, the fitting schedule issued for pricing.

You must lock: which substitutions are acceptable and on what evidence.

This is where schemes are most commonly destroyed, and the defence is written, not argued. Make the criteria measurable and mandatory: photometric distribution within a stated tolerance, Ra and R9 minima, SDCM initial and maintained, flicker limits (PstLM ≤ 1.0, SVM ≤ 0.4), driver compatibility with the specified protocol, physical aperture and depth, warranty terms, and full LM-80 data with TM-21 projection at the actual drive current and case temperature.

“Or approved equivalent” with nothing behind it is an invitation. A substitution should have to prove equivalence against numbers, and you should require samples of anything visible.

Construction

You produce: submittal reviews, RFI responses, mock-up reports, site inspection reports.

You must do: mock up anything where texture, finish or a specific effect is doing the work.

A full-size on-site mock-up with the real fitting, the real finish and the real geometry is the cheapest insurance in the process. No calculation predicts how a particular stone reads under a particular grazing angle, and no render is admissible evidence when the client disagrees.

Review every submittal against the actual specification clauses, not against the general impression that the product looks similar.

Focusing and handover

You produce: the aiming record, the commissioning sign-off, the O&M information, the final scene documentation.

You must have: this stage in the fee proposal, and the contractor’s obligation to provide access equipment and an electrician on standby, written into the contract.

Focusing happens at night, takes longer than anyone plans for, and is where the design is finally either delivered or lost. Every adjustable fitting is aimed, locked and recorded. Every scene is set with the client present. Every level is written down.

Then hand over documentation that lets someone else maintain it: what the scenes are for, what lamp and driver go in each fitting, what the CCT and SDCM tolerance is for replacements, who owns the control system, and when the tunable schedule should be reviewed.

The five failure points, in order of frequency

1. Ceiling void or cove detail decided without the lighting designer. Concept-stage failure, discovered in detailed design, unfixable.
2. Substitution at tender against a specification too vague to defend. Prevented only by writing measurable criteria.
3. Finishes changed after the calculations were signed off. Prevented by re-running against final reflectances.
4. Commissioning with no owner and no fee. Prevented by making it a named work package.
5. No focusing visit. Prevented by putting it in the fee proposal at the start, where the client can see and accept it.

None of the five is a technical problem. All five are process.