Keep the shell or demolish it? I want a rule, not a feeling.

Your instinct about the code is close, but I'd state the rule differently after twenty years of putting steel into other people's mistakes. Keep the shell when the structure gives you capacity you'd otherwise have to buy, and you can verify that capacity for less than it costs to replace.

On your building specifically. Bar joists at 30 foot spacing carrying roof only is the standard 1974 warehouse, and they're typically designed for something like 20 psf total roof load plus snow. That is a roof and nothing else. Any mezzanine, any second floor, any mechanical unit that wasn't in the original design, means new load path to new footings. Once you're setting new footings inside an existing building you're paying premium labor for restricted access, hand digging around unknown utilities, and shoring. Budget 1.6 to 2.0 times what the same footing costs in open air. That premium is what eats the shell's value.

The slab is the piece I'd chase before anything else. Five inches unreinforced on 1974 subgrade under a light industrial use is usually fine for residential live load, and usually terrible as a finished floor. Core it in eight or ten places, get subgrade samples, and get a vapor emission test. If you have to pour a topping slab and a new vapor barrier across 38,000 feet you've spent $8 to $14 a foot recovering something you thought you owned. That single test costs a few thousand and moves your Scheme A number by half a million either way.

What the shell actually buys you here: footings, the slab if it tests clean, and 16 foot clear that code will never hand you cheaply in stick framing. What it costs you: one story on a site zoned for three. You're giving up 36 units to save 12 months. On $273k versus $242k a unit that trade only works if the ceiling premium is real and durable.

My rule, if you want it as a sentence. If the existing structure can carry the new use without new vertical load path, keep it. If it can't, you're buying a very expensive site fence.

The number I'd stress test is the 8% ceiling premium. You called it a guess. On 26 units at, say, $1,600 base that's $128 a unit a month, $40k a year, roughly $600k of value at a 6.5 cap. That's most of the gap between the two schemes on a per unit basis. If it's actually 3% the argument for A gets thin fast.

Is there real leased product in your market with 14 foot plus ceilings you can pull actual rents from, or is the premium coming from listing photos?

Dumb question from the back. Why is demolition only $290k on 38,000 feet if there's asbestos in it? I thought abatement was the thing that blows budgets.

@ember it's not a dumb question and $290k reads low to me too. Metal panel and CMU comes down cheap, but suspect pipe insulation means a survey first, and if the survey finds friable material in more places than the estimator assumed the number moves in ugly steps. I had a 14,000 foot shop where the number went from $60k to $184k because of floor mastic nobody had bothered to test. Arbor, has anyone done an actual asbestos and lead survey or is that a phone quote off a site visit?

Something I don't understand well enough to argue with anyone about: does the reassessment work differently between the two? You're going from an industrial building on a low assessment to either 26 or 62 residential units. Scheme B's tax bill has to be much bigger in absolute terms, and it hits after a longer construction period. Does that show up anywhere in the $242k a unit, or is that hard cost only? Assessment practice varies a lot by state, which is the part I always forget.

Reading this three times and learning a lot. The thing that surprised me is that the more expensive project is the cheaper one per unit.

@rigging that's the piece I'd underline too. Fixed costs like the site work, the utility connections and the entitlement don't double when the unit count doubles, so they spread thinner. Arbor's 62 unit scheme is carrying the same driveway and the same sewer tap as the 26.

18 months versus 30 is not a small thing though. Two more winters of carry, two more years of construction pricing you can't lock, and a lease-up landing in a market nobody can see from here. I'd want to know what the loan looks like in each case before I called B the better project. A $7.1M project and a $15.0M project are different animals with different lenders and different guarantee requirements, and you should get that in writing from whoever is quoting you.

Can you do both? Phase it. Convert the shell now for 26 units on the front two acres and hold the back two for the stick build later.

@plumb that sounds good until you manage it. A 26 unit single story loft property is a different operating animal than a 62 unit three story, and if you build the second phase next to occupied units you're running a construction site through your renewal season. I've managed through that twice and lost tenants both times.

Arbor, the other operating note on Scheme A: single story with light courts means a long walk from parking to some units and a roof over every square foot of rentable area. Your per unit roof and mechanical maintenance is roughly double what it'd be stacked three high. That's not a deal killer, it's a line in the expense ratio that most conversion pro formas copy from a stacked comp and never adjust.

1,050 gross for a loft unit and 26 of them. What's the mix? If those are all one bedrooms at that size you're paying to heat space nobody's paying extra for. My small buildings have taught me that the sweet spot moves with the local renter, and in a small downtown the demand is usually thinner at the top of the size range than the architect assumes.

I'd underwrite the two schemes to different exits rather than the same cap rate, which is where most of these comparisons go wrong. The mistake is assuming a buyer prices both the same.

26 unit single story loft conversion is a specialty asset. Fewer buyers, more of them local, and the debt available to them is worse. 62 units of conventional stick over slab is agency-financeable product that a regional syndicator can bid on. That's usually 50 to 100 basis points of cap rate difference at exit, and on your NOI that's real money. Run B at 6.0 and A at 6.75 and see whether B's advantage widens or the ceiling premium holds it together.

Also @girder's slab coring point is the cheapest information in this thread by a wide margin. Do that before anything else.

Following mostly to learn, but one question. If Scheme B is better on every number except time, what's the actual reason A is still on your desk? Sometimes the answer is that you already own the building and it feels wasteful to knock it down. That's a real feeling and a bad reason.

On the mechanical side, single story with 16 foot ceilings and 26 units means 26 separate systems and a lot of roof penetrations, or a long horizontal distribution run and a big central plant. Neither is cheap. Whoever gave you $148 a foot, ask what they assumed for HVAC and whether they'd stand behind it after a real MEP drawing set. My experience is that the first number on conversions always assumes the easy version of the mechanical.