Report 078 · Luxury Home Security
The sticker on a nine-thousand-dollar safe says UL. It does not say which of the three relevant standards, or which tier inside that standard, and those are the only two facts that describe what you bought. Until 2016 the residential burglary test was a single five-minute attack by one person with hand tools.
A safe is the last layer. Everything else in a residential security design is meant to make an intruder leave before they get to it: the cameras, the alarm, the lighting, the response. The safe is what protects the contents once all of that has already failed. Which makes it the one component where the specification deserves to be read rather than trusted.
And it is the component where the marketing has drifted furthest from the document. "UL rated," "UL listed," "fireproof," "burglary rated," all appear on product pages without saying which standard, which class, or which tier. Three different UL standards can be behind that word, they test unrelated failure modes, and one of them explicitly declines to tell you whether the safe still works afterward.
All the documents quoted below are the free, publicly published scope text from UL and UL Standards & Engagement. None of it required a purchase, and none of it appears in the marketing.
The rating on most home safes and effectively all gun safes is Residential Security Container, and it does not live in a burglary standard at all. UL's own announcement puts it plainly: "RSCs are covered by UL 1037, the Standard for Antitheft Alarms and Devices."
Here is the part worth knowing. For decades the RSC rating was a single pass/fail:
"Prior to the change, there was a one-size-fits-all level of performance for RSCs: a five-minute attack test using common hand tools."
Five minutes. One test. Everything from a two-hundred-dollar closet box to a two-ton gun safe earned the identical designation.
The Sixth Edition, revised in 2016, split that into three attack levels. In UL's words:
Notice what varies between the tiers. Not just the clock: the number of attackers, the tool set, and the definition of what counts as a breach. Level 3 is harder partly because the opening the attackers are allowed to make is a third the size, which is a much stricter definition of failure than "they got in."
There is also a mounting clause that almost nobody quotes and that changes how you install the thing: products under 750 pounds must pass a drop test simulating forcible impact, unless the manufacturer supplies reliable mounting provisions and installation instructions. The lighter safe is certified on the assumption that it is bolted down. A 600-pound safe standing free on a closet floor is outside the condition its own rating was granted under.
The practical consequence of the 2016 split is a market where all three tiers, plus a large population of pre-2016 units, carry labels that read similarly. "UL RSC" without a level is, at best, Level 1: five minutes, one person, a drill and a hammer. That is a real amount of protection against an opportunist and close to none against anyone who came prepared, which is a different question I looked at when writing about how burglars actually case a luxury home.
The classes people associate with serious safes, TL-15 and TL-30, are not in UL 1037 at all. They come from UL 687, Burglary-Resistant Safes, and the free scope text lists them in a structure that tells you more than the names do:
"These requirements cover combination locked burglary-resistant safes classified as follows: a) Test attack against the door and front face: 1) Tool-Resistant Safe – Class TL-15 2) Tool-Resistant Safe – Deposit Safe 3) Tool-Resistant Safe – Class TL-30 4) Torch- and Tool-Resistant Safe – Class TRTL-30 b) Test attack against the door and body: 1) Tool-Resistant Safe – Class TL-15X6 2) Tool-Resistant Safe – Class TL-30X6 3) Torch- and Tool-Resistant Safe – Class TRTL-15X6 4) Torch- and Tool-Resistant Safe – Class TRTL-30X6 5) Torch- and Tool-Resistant Safe – Class TRTL-60X6 6) Torch-, Explosive-, and Tool-Resistant Safe – Class TXTL-60X6"
The organizing principle is not the number. It is the two-part split at the top: some classes are tested by attacking the door and front face only, and others by attacking the door and the body. That is the X6 suffix. A TL-30 has been attacked at the front. A TL-30X6 has been attacked everywhere, which is where an unhurried thief with a floor and a wall available would actually go.
TR means torch resistance is also tested. TX adds explosives. Those are qualitative jumps, not increments.
And here is an honest limit on what I can tell you. The classification list is free. What the numbers denote in test terms, the net working time and the tool schedule, sits inside the purchasable standard. Every dealer page will tell you TL-15 means fifteen minutes of net working time by a trained attacker, and that reading is consistent and probably right, but it is a dealer's gloss and I have not opened the clause. I am flagging it rather than repeating it as fact, because the point of this report is that the buyer-facing meaning of these labels is assembled almost entirely from vendor copy.
Fire resistance is UL 72, Tests for Fire Resistance of Record Protection Equipment, and its published scope is the most useful paragraph in this whole subject:
"These requirements cover the test procedures applicable to the fire-resistance classification of record protection equipment intended to provide protection to one or more types of records when exposed to various durations of fire exposure."
Records. The standard is not about protecting valuables from fire in general, it is about protecting records, and its classes are defined by what kind of record is inside:
The class number is the interior temperature ceiling in Fahrenheit, which is why the ordering runs backwards from intuition: Class 125 is the most demanding, not the least. Paper chars at a high temperature. Magnetic media and film give up far earlier, and disks earlier still.
The structural giveaway is in the description of the fire endurance test, which is written to demonstrate "the ability of the device to prevent the development of temperatures and relative humidity (Classes 125 and 150 only) exceeding the specified limits inside the device."
Classes 125 and 150 only. Humidity is not measured in a Class 350 test, because paper tolerates the steam that comes off insulation during a fire and a tape reel does not. So a paper-rated safe can pass its test in a condition that would ruin the contents of a media-rated one, and the certification is still perfectly valid, because it certified a different promise.
This is the mismatch that matters in a house. A Class 350 one-hour safe is an entirely appropriate place for deeds, passports and a will. It is not a rated enclosure for backup drives, negatives, or anything with a circuit board in it, and no amount of "fireproof" on the box changes what was measured.
Section 1.3 of the same scope, verbatim:
"Tests conducted in accordance with these requirements are intended to demonstrate the performance of record protection equipment during exposure to fire, but are not intended to determine acceptability for use after fire exposure."
The certification covers the contents, during the fire. It makes no claim that the safe is still a safe afterward. The lock may be destroyed, the seal spent, the body warped. That is not a defect, it is outside the scope of the test, and UL says so in the free text.
Which means the recovery plan matters as much as the rating. If the fire is the event you are protecting against, plan on the safe being a single-use container that survived long enough, and on needing a locksmith to open it.
Two more things in that scope are worth carrying. The fire and impact test requires the device "to protect contents from heat, to the extent described in the requirements, before and after an impact due to falling 30 ft," which models a safe on an upper floor going through the floor when the structure fails. And there is a separate explosion test, meant to show the device can withstand sudden exposure to high temperatures without the contained steam or gases blowing it apart. A safe put into a fire cold and fast is being asked a different question than one heated gradually.
Six questions, all answerable from a spec sheet or a phone call, none of them answerable from the word "UL":
That last one is the one people skip, and it is the only one that converts a rating into a plan. A five-minute container is genuinely adequate if help is four minutes out and genuinely decorative if nobody is coming, which is not a hypothetical: it is the situation across a growing list of cities, as I wrote a few days ago in the report on verified-response ordinances.
This is the part of the problem I work on. Detection and response timing is what determines whether a delay rating is a real defense or a decoration, and getting the alert right (fast, verified, and specific enough to act on) is the difference. I help design the AI security systems for a veteran-owned (SDVOSB) home-security company run by fellow veterans. I do not own it and earn nothing from this link; I flag it because it is a field I build in and not only write about. Full policy here.
Every certification in this article is a stopwatch, and every one of them is honest about it. Five minutes, one person, hand tools. Ten minutes, two people, carbide drills, a two-square-inch opening. One hour below a stated interior temperature, humidity measured only if the contents need it, no claim about the state of the safe afterward.
The documents are precise. The labels are not, because "UL rated" compresses three unrelated standards and, within one of them, three tiers separated by a factor of several in difficulty, into two characters.
Read the standard number and the class. Then ask how long until someone arrives. A safe does not stop an intruder. It buys time, and time is only worth something if it is being spent on the way to somewhere.
Scope note: the quoted material is the free, publicly published scope and announcement text from UL and UL Standards & Engagement. The full standards are purchasable documents and were not obtained, so this report does not state the net working times, tool schedules or test temperatures held inside them; where a widely repeated reading of a class name (such as TL-15 denoting fifteen minutes of net working time) comes from dealer sources rather than the standard, it is identified as such in the text and not asserted. The statement that a class number in UL 72 corresponds to the interior temperature ceiling in degrees Fahrenheit is the conventional meaning of the designation and is consistent with UL's published class definitions, but the numeric limits themselves are inside the purchasable standard. Certification status and edition vary by product; check the actual label and the manufacturer's listing rather than assuming a rating from a product category. This is general security analysis, not a recommendation about any specific product or an assessment of any specific property.
U.S. Army combat veteran (Counter-IED / Electronic Warfare), peer-reviewed researcher in microwave spectroscopy, and founder & CEO of Shroombiosis. Consults on laboratory operations, AI, and supplement formulation.