Hello,

I work for an company which we mandate 0.3 insulation factor and never go below... in API521 there is allowance to go lower..but this will massivly reduce the relief valve size.. so from a cost perspective very attrative but safety wise..relys on high integrity/or robust management of insulation which to me seems a step to far..

However, api 521 does allow for this.. so i thought i would challenge my mindset and findout if other companys actually use lower factors..
 

Heavy, well managed insulation with a aluminum or stainless facing, plus a "water curtain" from a high-reliability Firewater system and pre-aimed nozzle would let you go lower.  

Now, tell me if you have ever seen a setup that robust.  I havn't.  I stop at 0.4 - seems to me to be the realistic low-limit.

Thats a fair point.. API allows 0.3 per the standard.  and ive seen that at a few companies.. i know API does not allo wa water curtain or allowance for firewater credit..

however the the old rospa codes do.. i dont know of any other codes that allow fire water credit..
 

Yes, I use the API fire heat input calculations, and I use the resulting F value, which is typically much lower then 0.3. In my experience, that's common industry practice. In the past, I've informally surveyed other companies (5-6 large companies) and found a range of responses. I found that some engineering companies will use a min value of 0.3 by default, unless specifically authorized by the client to go lower. I found one operating company that limited the value to 0.15, and one that had a limit of 0.05. The others had no company mandated limit.

Personally, I'm OK with the low F values as long as you ensure that you're using a fire-resistant insulation material and SS covering.

The 0.3 value is the NFPA-30 credit value. Some companies calculate the heat input for pressure vessels based on NFPA-30, but most use the method in API 521.

The driver for me is how much influence the issue has on the flare system design and the stack height.  If you are not constrained by that, then there really isn't much of a driver to go lower than 0.3.  However, in pre-existing plants undergoing upgrades or debottleneck projects, there may indeed be a very strong driver to take more credit for insulation in order to avoid a massive rebuild of the flare system.

Regards,

SNORGY.

so does this mean that for my case with a completely uninsulated (bare steel) heat exchanger that is in doors, I should use a factor more like one? There are no immediate fire hazards within 20-25 feet of the exchanger, but still... Please let me know what you think!

What is the size for the relief valve in the fire case for the bare steel unit, and what are the associated costs in the relief valve and flare system, versus the savings achievable in these things by insulation / fireproofing?

well, i'm actually trying to calculate the necessary size for a PRV to stay within code requirements and the program I have is asking for an environment factor which I'm assuming is what you are talking about in this thread. Please correct me if I'm wrong. The current PRV on the exchanger has an area of 0.11 square inches and the set point is at 175 psig, so it can flow about 38 gpm. I'm trying to figure out if this is the correct size or not since we need to replace it (it's leaking) and I just want to verify it's correct rather than hoping.

0.11 in2 is a D orifice which is the smallest API orifice PSV you'll get (there are smaller orificed PSVs though). I'd try the environmental factor set at 1.0 and if your 0.11 in2 PSV is big enough, the easiest replacement is just to replace it in kind.

In my refinery experience all the big vessels are protected with API 521 insulation to obtain F=0.3 (hundreds of cases).
This kind of insulation is hated by the construction teams: very difficult to be applied, certification not easy to be obtained.
F lower than 0.3 are only applied (only one project, in a gas field) where forced by restrictive max allowable flare loads. In these cases the insulation is provided at the maximum extent but the PSVs are sized in a more conservative way (anyway within the flare load limits).