Will Your Exhibition Display Meet the Venue’s Fire Safety Rules?

Methodology, scope and limitations | Stand Central original research

Research question

What are UK exhibition stand items made of, what do supplier product pages say about their fire performance, and what fire and safety rules do UK venues and organisers apply to them? The study reports what pages and rule documents state. It makes no statement that any product is or is not compliant, and it does not assess whether one fire standard is equivalent to another.

Sample

We read 1,034 product pages from 39 UK display suppliers, the exhibitor rules of 15 UK venues, 20 organiser exhibitor manuals and the AEV eGuide (August 2026 edition). Product pages come from 778 specialist display supplier pages and 256 online printer pages. The data hold 1,537 coded rows: 1,034 product-page rows and 503 rule-variable rows from 36 rule sets (the eGuide, 15 venues and 20 organiser manuals).

Suppliers, venues and organisers from which no usable data was obtained are excluded from the study and are not described as part of the sample. Stand Central, which commissioned the study, was excluded from the sample before collection.

The organiser exhibitor manuals are a purposive context sample of current 2026 and 2027 UK show manuals found on show websites. They are analysed separately from venues and are not presented as representative of any wider set of shows.

Timeline and protocol

A timestamped protocol was written at 2026-09-29T11:21:29Z, before any product page or rule document was collected for coding. It defines the questions, units, sampling frames, eligibility, missing-data rule and analysis. Every later change to the method was logged with its date and time. The changes to collection and coding were all made before any share or tally was computed. Three changes came after the first analysis and are disclosed here. First, material codes taken from link text pointing to other products were removed (no fire, standard or certificate value changed). Second, suppliers, venues and organisers that gave no usable data are not reported as part of the sample, which changed no computed value. Third, the same rule was applied to venues and organisers whose captured documents gave no usable rule data, so they left the sample; no value of any retained rule set or product page changed.

All sources were captured between 2026-09-29T11:19:28Z and 2026-09-29T12:15:38Z (UTC). Results describe pages and documents at that time.

Product pages (Part A)

Unit: one product page of one item type from one supplier. Product URLs were taken from each supplier’s sitemap (or home-page links), classified into 12 item types by fixed keyword rules, and sampled in a seeded random order: up to 4 pages per supplier and item type. Listing pages, accessories and spare parts, marketplace listings and duplicate content were excluded by fixed rules. Pages that failed a plain fetch had one headless-browser attempt; human-verification pages were never bypassed.

Coding: a deterministic rule-based coder searched the full captured text of each page (template lines repeated across a supplier’s site and related-product blocks removed) and stored the exact matching line as a verbatim excerpt with a line locator. Fire statement categories: names a fire test standard; fire-retardant wording with no standard; states the product is not fire rated; fire wording only in site-wide template text; no fire statement. A bare “B1” in a fire context is counted as a DIN 4102 B1 claim and reported separately. Materials are coded multi-label, only where the product text names them.

Venue and organiser rules (Part B)

Unit: one rule set, meaning the current published exhibitor, technical or event regulations of one venue, or one organiser manual for one show edition. Several documents of one venue are pooled, and each value keeps its own source. Venue values are coded from the venue’s own documents; a copy of the eGuide hosted or linked by the venue is recorded separately. The AEV eGuide August 2026 (Stand Build and Electrical sub-sections) is coded as the reference rule set.

Fourteen rule variables were coded with verbatim excerpts, including the fire standard named for stand materials, whether test certificates must be available, on-site testing, height triggers, working at height, risk assessments, CDM, electrical testing, bans, stated consequences and how the rules refer to the eGuide. A silent rule set is coded not_stated; silence is never read as permission. 7 coded rows carry an analyst override, each with a stated reason in the data.

eGuide reproduction test: 11 key materials clauses fixed from the August 2026 Stand Build sub-section. A rule set reproduces a clause when any 12 consecutive normalised words of the clause appear in its text (lower case, punctuation removed, and the letter pairs fl and fi removed from both texts because some PDF text layers drop those ligatures).

How each central finding is calculated

Silent pages: product pages with no fire statement or only site-wide template wording, divided by all 1,034 product pages = 824 / 1034 = 79.7%. Indoor-only share: 79.9% of 988 pages.

Standards: of 144 pages naming a standard, pages naming any DIN 4102 code or a bare B1 = 119 (82.6%); naming BS 476-7 = 4. eGuide mentions: regex line counts over the full August 2026 eGuide text (BS 476 on 4 lines; DIN 4102, B1, EN 13501 and BS 5867 on 0 lines).

Item types: per item type, pages whose title confirms the type, stating fire performance (named standard or fire-retardant wording) divided by those pages. Table covers 44/67, fabric backwalls 22/108, counters and plinths 7/101.

Certificates: product pages offering a fire certificate to download or on request = 10 of 1,034 (1.0%), from 3 suppliers. Rules restating the certificate rule in their own text: 1 of 15 venues and 3 of 20 organiser manuals. Venues pointing to the eGuide: own-text reference to the eGuide or a hosted copy = 11 of 15.

Verification

Every coded value keeps its source ID, URL, publisher, UTC capture time and verbatim excerpt with a locator. Every excerpt is checked by script against the saved capture. All derived numbers are regenerated by script from the observations, and the calculation behind each finding is set out above.

Independent live re-check on 29 September 2026: 1,081 of 1,081 sources behind the claims were fetched again. All excerpts were found again on 856 of 856 rows with excerpts, and 824 of 824 rows coded as having no product-specific fire statement still showed no fire term on the fresh page. 0 rows were unavailable.

This publication package recomputes every headline number independently from the canonical observations and checks it against the canonical analysis before any file is written.

Limitations

Absence is page-level. A product page with no fire statement does not mean the product fails a fire test or lacks a certificate; it means the buyer cannot tell from the page.

Certificate offers are coded from page text, not from privately held paperwork. A certificate file linked without any certificate wording on the page would be missed.

Item-type comparisons measure published information, not fire performance. We did not test materials, observe stand acceptance or measure why suppliers describe categories differently. Statement counts do not establish fabric area, treatment, printing or finished construction.

Material shares are over pages that name a material; pages that name none stay unknown, not zero.

The product sample is seeded and random within each supplier and item type. It is not weighted by sales or market share.

Part B coding is rule-based with a small number of documented analyst overrides. Two PDF text layers lost the fl and fi ligatures; the verbatim test handles this symmetrically.

One supplier’s site placed part of its catalogue behind human verification, which was not bypassed; its captured pages are included.

The study does not assess compliance, the equivalence of different fire standards, or what any venue will accept on site. Venues, organisers and their safety officers decide that.

Findings describe pages and documents at capture time (29 September 2026) and may change.

Commercial-interest disclosure

Commissioned by Stand Central, a UK retailer of exhibition displays. Stand Central was excluded from the sample before collection and is not coded, counted or named in the data. Findings describe what pages and rule documents state, not product quality or safety.

Data and reproduction

The accompanying CSV files hold every product-page and rule-variable row with public field names, a rule-set summary, the item-type table and a source register with minimal verbatim excerpts, locators, publishers, URLs and capture times. Raw captures and internal paths are not distributed. Blank cells are null or empty in the canonical data, never zero.

Primary source links

AEV (eGuide) (EG-2026-FULL) | https://cdn.asp.events/CLIENT_AEV_30A5AAAF_5056_B740_1746BDBB4A667186/sites/aev2022/media/eGuide/2026/eGuide-August-2026-clean-version.pdf

AEV (eGuide) (EG-2026-STANDBUILD) | https://cdn.asp.events/CLIENT_AEV_30A5AAAF_5056_B740_1746BDBB4A667186/sites/aev2022/media/eGuide/2026/2026-subsections/Stand-Build-August-2026-subsection.pdf

EAGE (organiser) (R-PJ-EAGE) | https://eageannual.org/wp-content/uploads/sites/11/2026/04/EAGE-Exhibiting-at-PJ-Live-A-Z-Document-v2.pdf

Excel London (R-EXCEL-RIG) | https://www.excel.london/uploads/rigging-guidelines-and-checklist.pdf

Floating Offshore Wind (RenewableUK) (O-FOW-AZ) | https://exhibitormanuals.co.uk/fow-a-z-2026

JTrade (O-JT-MAN) | https://jtrade.co.uk/wp-content/uploads/2026/03/JTrade-2026-Exhibitor-Manual.pdf-4.pdf
