The Fireproof Cat

Revision history

This study has been revised three times in eleven months. Each revision was prompted by the Client's rejection of the one before it. The scope has widened on each occasion. It is no longer narrowing.

What has been tried, and by whom.

The Division is not the first body to attempt the fire-rating of a living animal. The literature is thin, frequently withdrawn, and in one case the Division is not permitted to discuss the reasons for its withdrawal. We cite it anyway. It remains the most directly applicable prior work on the record.

1. Field trials of a fire-resistant garment for relocated ungulates, HIB-MAT-2023-0022, Hadrian Institute, 2023. Conducted on wildebeest in holding. Abandoned at Phase 2 after the test article shed the garment, cleared a 1.4 m fence, and was recovered 11 km away in a neighbouring host property. The garment was recovered separately. Findings carried forward where applicable; the wildebeest was not.
2. V. Albright & S. Okafor, “Autoignition characteristics of mammalian α-keratin fibres under domestic heat loads,” J. Appl. Fire Sci., 2021. The definitive study of why fur burns. The Division has read it carefully and has nothing to add.
3. A Review of Spontaneous Combustion in Domestic Companion Animals, author redacted, 2019. Concludes the phenomenon is “statistically rare and largely anecdotal,” which the Division considers precisely the correct amount of reassurance to offer the Client.
4. Materials considerations for elevated ungulate housing, HIB-CIV-2025-0093, Hadrian Institute, 2025. The Nuneaton giraffe-tower feasibility study. Cited for its treatment of live-load behaviour in biological structures, which the present study found, unexpectedly, relevant.

The cat is combustible because it is alive.

The Client requires a domestic companion animal that does not constitute an ignition hazard under foreseeable household conditions. This study finds that the requirement cannot be met in full. The cat is combustible not by defect but by composition. Every subsystem that contributes to cat-hood — a keratin integument, a lipid reserve, a glucose metabolism, a set of keratin sensors — is a carbon-based fuel, arranged over thirty million years into a configuration optimised for warmth, mobility, and a slow, sustained burn.

We further find that fireproofing and cat-hood exist in direct and unrecoverable opposition. Each increment of fire resistance removes one increment of cat. The relationship is monotonic. It does not reverse. The point at which the animal stops being a cat is a matter for the Client, not for this Division. We cannot select it on the Client's behalf, and we have declined to.

Pending the Client's selection of an operating point on the spectrum at §6, we recommend the interim measure at §7.1. It is the only recommendation in this report that does not require the Client to decide what a cat is.

What the Client asked for.

The Client underwrites household companion-animal liability across the United Kingdom. At present, the fire risk posed by the domestic cat is excluded from every standard policy. The exclusion is broad, it is not well understood by policyholders, and it is the single largest source of complaint received by the Client's claims department. The Client wishes to remove the exclusion. To do so it requires a cat that does not, itself, remove the exclusion.

The brief, as received, is reproduced below in the Client's words. It has not been edited for register.

“Make the cat out of materials that do not catch fire.” — Instruction HFDU-26-118, received 09 July 2025

For the avoidance of doubt, the Division has interpreted the instruction as the following acceptance criteria. The animal, post-modification, shall:

1. not sustain combustion following exposure to a domestic ignition source (candle, gas hob, log fire) for a period not exceeding three (3) seconds;
2. retain the ability to locomote, to ascend to elevated surfaces, and to displace objects from them;
3. retain the ability to produce a sustained low-frequency vibration, hereafter “purring”;
4. remain, in the reasonable judgement of a policyholder, a cat.

Criteria 1 through 3 are engineering problems. Criterion 4 is not. It is the criterion against which every prior revision has failed, and it is the criterion the Client has, to date, declined to define.

The cat, as supplied.

The baseline animal — designated Test Article TA-3 for the duration of the trial — was a neutered male domestic shorthair, 4.2 kg, approximately four years old, loaned to the Division by the Client from the policyholder against whose claim the original exclusion was first invoked. TA-3 was, throughout the audit, uncooperative. The Division records this as an observation, not as a finding.

Six primary subsystems were examined against the acceptance criteria. The audited layers correspond to the cross-section at Figure 2. The results are below.

Of six audited subsystems, one — the skeleton — is fire-rated. The remaining five constitute, between them, a self-fuelling ignition event held in check only by the animal's continued preference for not being on fire. This preference is reliable but undocumented, and the Division does not consider it a suitable basis on which to underwrite a policy.

What we advise.

In summary: the Client's requirement has, over eleven months and three revisions, not been met, and is not, in the Division's assessment, meetable. The cat cannot be made out of materials that do not catch fire without ceasing to be a cat. A garment can. We recommend the garment.