The Ridiculous Idea of a Hardened Shelter

So for years we’ve successfully designed severe storm (tornado) shelters in the Midwest USA using either FEMA 361 or ICC 500 as the standard for design.

These standards provide guidance on wind speed, design criteria, detailing recommendations, projectile resistance, venting, etc. This provides the engineer with a standard to base the design on that has been peer reviewed and adopted by the building codes.

Recently we’ve seen a lot of clients ask for a “hardened” shelter, but not a FEMA shelter. This, I believe, is their attempt to save a little money by having a “safer” space than normal without the cost of a full blown ICC 500 shelter.

The concern I always have is that when the A/E team obliges them, and designs a “sort of” FEMA shelter, the owner then subsequently plasters a sign on the area calling it a “shelter” and occupants rush into the space thinking they are safe.

A tornado comes along and a 10 ft. 2x4 pierces the non-FEMA (i.e. hollow metal doors) and someone gets injured or killed. Then the family sues the owner, and/or the owner sues the A/E team saying that they didn’t “harden” it adequately.

Since we can’t design to a non-existing standard (there’s no such thing in the A/E world as “hardened”) then the A/E is at an elevated risk. The word “Hardened” has no real meaning or measure.

So what to do? Some type of written communication is needed back to the owner stating exactly what we are, and aren’t, designing for. Not sure it would avoid a lawsuit but would certainly help in our defense.

Have any of your experienced this? What has been your response?


Never dealt with FEMA storm shelters, but have dabbled with different tsunami shelter configurations. Nothing serious though.

I have walked away from projects though where the client wanted me to bend the code, increasing my liability. Done that several times and never looked back.

You are getting into a new, undefined realm that requires caution. Would your liability insurer want you to go there? Probably not.

Being me, I would walk away until the dust settles and you can see clearly.

I think I know the answer (and I’ve been in a similar position myself I can add) but here’s a question. I think this warrants a discussion that includes some of the many ways this issue could be resolved. There is nothing inherently wrong with a client asking for an additional feature, improvement, or, by extrapolation, aesthetic flair.

What if you acceded to the client’s request to “harden” the door (rivet a steel sheet on to it, for the sake of argument) and did something you consider reasonable, but wrote explicitly in documents that you review with the client that there is no way to guarantee how much protection is added?

I can think of a few things both for and against modifying this storm shelter’s door, but I’m not a civvy so I don’t know how you would prioritize them.

I agree;

  • A contract mod that stipulates that the design is not to FEMA standards.
  • Each drawing should likewise have the same caveat
  • The customer should sign off, in writing, whatever design reviews they are privy to, which should all state the same caveat

Ok, but no contract will ever protect you from being dragged into court by an unreasonable idiot.

Only walking away will do that.

Well we can’t walk away from a long term client (architect).
But we are, as you suggest, going to visit with our insurance rep. The problem is, this topic is a bit subtle with respect to standards and codes so not sure how much an insurance person could understand the nuances.

We have drafted a letter to be sent up the chain to clarify to what level of “hardness” we are and are not designing to. Whether that will do any good I’m not sure.

In terms of wind speed, we can certainly define a slower wind (FEMA 361 provides various design wind speeds vs. tornado intensity as well as the percentage “chance” of how often a certain level of tornado occurs.). That is the easy part of this.

But the projectile resistance is only based on various lab tests over the years and correlating those to a slower wind speed is impossible. The standard test for tornadoes is a 15 lb. 2x4 shot into a wall/door at 100 mph. Where that criteria comes from with respect to our typical 250 mph tornado speed is unknown. So if we were to design to a slower wind - say 150 mph - I’m not sure what sort of projectile criteria we’d even begin to use.

Also - there are numerous videos out there showing how even lighter/slower projectiles can easily penetrate a standard hollow metal door.

Per my original post - occupants will rush into this “hardened” space thinking they are safe. Well, maybe safer. But not necessarily safe.


OK. I get it.

Hopefully you have had a heart to heart with the Architect then so they do not make any misleading claims to their client…


Beginning your design process with the lawyers and insurance agents is not an auspicious start.

There’s something subtle that I’m wondering about… Why would you even consider testing your “addition”? What if they don’t want you to test it?

I’m not asking this stupid question to be facetious - I’m giving the architect the benefit of the doubt that maybe they are exploring what the choices and options are, and did not have a specific criterion in mind, aside that one that’s out of the end user’s price range. If they are exploratory questions, not a specific demand, then you may still have room to explore with the architect and the end user just what they should and shouldn’t expect. This kind of discussion would serve to clarify the realm of the possible in the minds of the end users as well as the realm of what you can realistically offer, all at the same time.

Assuming of course that this is being discussed with folks that operate that way…
And like Msquared48 has observed, having a middle-man in between makes any nuanced communication harder.

Hi SparWeb!
No - we wouldn’t ever do any testing - we have no equipment/lab for things like that.

I was referring to tests already published. Texas Tech has done a lot of them but there really isn’t a correlation between their tests (15 lbs. x 100 mph) and our ICC 500 wind speeds. Basically you either have a projectile resistant envelope, or you don’t.

The mandate to design to a “hardened” area is something that is essentially just given to us up front when we are providing a scope/fee proposal to our architect clients. We have no say in the preliminary decision to do this. They just say - “the client wants a portion of the building to have a hardened space but doesn’t want to have a full FEMA shelter designed”.

So we can either try to convince the owner that they should simply do a FEMA shelter, or we design to something lower (hardened) and then try to cover our butts by communicating to everyone the possible limits of protection that this less-than-FEMA area provides.

Oh. Good grief.
“Extrapolating” one test result to another condition is often hard, if not impossible, especially when you haven’t the means to test any of them. What you describe sounds like quite the bridge to cross.
The level of protection that you can assure without meeting the definite standard is “none”, isn’t it?

None that are provable, anyway.

So, it has to be crystal clear legally and conversationally to the client, they shoulder all the risk for any deviation from the standard.

ISTR that the 100mph 2x4 was a testable compromise that was added to the Miami Dade building code shortly after Hurricane Andrew.

That would take one big slingshot!

Well that’s simple.
Test your non-standard hardened door with a 2x4 at 90 mph instead.

Sparweb - yes - without a fixed standard for a lower than-FEMA shelter (i.e. a hardended shelter) there is no level of protection that we can provide… Agree 100%. That’s why this whole idea of our architects allowing their clients to push them into these “hardened” type spaces is, as I said in the title of this, ridiculous.

Is it worth your time to establish a project-specific standard (or perhaps a standard for your company assuming this request begins to repeat) for what this “hardened” shelter will look like? You could pull in multiple resources from various other industries to build whatever design criteria you deem appropriate - for example, the NRC has a Regulatory Guide that establishes design criteria for both tornado missiles and hurricane missiles for which nuclear power plants must withstand - I’m not suggesting that you design your shelter like a nuclear plant, but perhaps borrowing some of their empirical equations for penetration, as an example, may at least give you a basis to design to. I wouldn’t doubt that other resources exist outside of FEMA where more information is available similar to that instance.

I know it would be a significant effort in terms of time to build such a standard - perhaps your architect client would go in with you in the endeavor?

Edit: The Regulatory Guides I mentioned above are RG 1.76 and RG 1.221. These guides are readily available from a search of the NRC’s website:

Option # 1
Ask the architect to specify exactly what enhancements to a standard structure he considers “Hardened”.
ie: Have the architect provide the definition of “hardened”.
Option # 2
A letter worded more diplomatically that this sent to the architect:
Before attempting to comply with your request, we would have to involve our lawyers and insurance representatives in discussions with you to determine and avoid possible liabilities.
We are seriously considering that a project started with lawyers and insurers may be better not started.
Following re-assessment of possible liabilities, you may agree with this suggestion.
Option # 3
Have the owner and or architects sign off after construction is completed that they have inspected the enhancements and determined that said enhancements meet with their approval.
Furthermore, they have done due diligence and inspected the enhancements to the basic code and accept that the design, materials and workmanship meet their expectations unreservedly.
They accept the enhancements as is and where is and assume all responsibility for any future liability arising out of any use or misuse of the space.
I like option # 2 the best.

1 Like

KoachC SR - I’ll look at the nuke site and see what I can glean from it.

WaRoss: I have already written a letter just to see how it looks to me. I’ll post it here in a minute. Your Option 2 is what I would / will do with my architect(s) on the phone - just trying to convey to them that they are in the same boat as I am (the A/E team could be held liable for injuries to to a hardened space that wasn’t so hard after all). Option 1 - the architect would probably respond like a deer in the headlights. Option 3 - My insurance rep that I visited with yesterday suggested that some type of indemnification might be attempted but hard to get.

Here’s a sample letter that I drafted. What do you think?
Client letter for HARDENED spaces.pdf (23.6 KB)