Flat roof snow load vs minimum snow load: the two 0.7 factors that get conflated

ASCE 7-22 Chapter 7 produces two candidate uniform roof snow loads for a low-slope roof: the flat roof snow load p_f (and the sloped-roof p_s that follows from it), and the minimum snow load p_m. Both are strength-level loads. Both get run through the Chapter 2 load combinations. And with 7-22's move to strength-based ground snow loads, both workflows now contain a factor of 0.7 — two different factors of 0.7 that do entirely different jobs. STRUCTURE magazine's November 2025 treatment of the topic (Showalter & Adams) documents that designers have been conflating them in practice, and it's easy to see why.

The first 0.7: ground-to-roof conversion inside p_f

The flat roof snow load is built from the mapped ground snow load:

The 0.7 here is a ground-to-roof conversion factor. It reflects the physical observation that a typical roof retains less snow than the ground next to it — wind scours it, heat loss melts it, some slides off. It is part of computing the load itself, before any load combination is applied. C_e (exposure) and C_t (thermal) then adjust that conversion for the specific roof.

What changed in 7-22: the ground snow load p_g is now a strength-level, reliability-targeted value, mapped by risk category, rather than the old uniform-hazard 50-year value scaled by an importance factor. That editorial change is exactly what put the second 0.7 into the picture.

The second 0.7: strength-to-ASD in the load combinations

Because the 7-22 snow loads are strength-level, the allowable stress design combinations of Chapter 2 apply a 0.7 factor to S — analogous to the familiar 0.7E for earthquake loads. An ASD gravity-plus-snow check is:

This 0.7 is a strength-to-service conversion. It belongs to the load combination, not to the snow load. It applies exactly once, and it applies to whichever snow case is being checked — the balanced p_f-based case, drift, unbalanced, and the minimum load p_m.

Where p_m fits — and where it doesn't

The minimum snow load exists because on low-slope roofs in areas of modest ground load, a single storm can deposit more snow than the fully-reduced balanced design load. So the code sets a floor, taken from the ground snow load (with an upper cap), without the 0.7 · C_e · C_t conversion chain — a single storm doesn't give wind and heat loss time to do their work.

Two rules follow, and both are stated by the code and confirmed in the STRUCTURE article:

• p_m is a separate uniform load case. It is checked on its own, against the full roof. It is never combined with, or added to, drift, sliding, unbalanced, or partial snow loading. Those cases build from p_f/p_s; the minimum load stands alone.
• p_m gets the ASD 0.7 exactly once, from the combination. In ASD you check D + 0.7·p_m, the same way you check D + 0.7·p_s. What you must not do is "convert" p_m with an additional 0.7 on the reasoning that every other snow number got one — the 0.7 that p_f got was the ground-to-roof factor, which p_m deliberately does not receive.

The error, with numbers

Take a heated, normally exposed low-slope roof with C_e = 1.0, C_t = 1.0, and a mapped strength-level p_g of 20 psf, and suppose the tabulated minimum for the site works out to p_m = 20 psf.

The correct ASD demand for the minimum case:

The conflation error — applying the ground-to-roof 0.7 to p_m first, then the ASD 0.7 on top:

Run the same slip the other direction — skipping the ASD 0.7 because "p_m already looks unreduced" — and you overshoot instead. Either way the design load is wrong, and a checker who only sees the final combined pressure may not catch which 0.7 you meant.

When does the minimum actually govern?

Compare the two uniform cases. The minimum governs when p_m > 0.7 · C_e · C_t · p_g — equivalently, when the ground load is modest enough that the tabulated minimum outruns the converted flat-roof load. In practice that means low-ground-snow regions and any roof whose C_e · C_t chain reduces p_f further. In heavy snow country the balanced, drift and unbalanced cases pull ahead and p_m fades to a formality — but it still must be checked as its own case.

How StructuralClaw handles it

StructuralClaw's snow module computes the balanced, unbalanced, drift and sliding cases per ASCE 7-22 Chapter 7 with each factor cited, and the project's load-combination engine treats the minimum snow load as its own basic load case — factored exactly like S in every combination, one 0.7 in ASD, never stacked onto drift or unbalanced loading. That separation isn't a convenience; it is what the code requires, and building it into the case structure means the error described above has nowhere to happen.

• ASCE/SEI 7-22, Minimum Design Loads and Associated Criteria for Buildings and Other Structures, Chapter 7 (snow loads) and Chapter 2 (load combinations).
• Showalter & Adams, "ASCE 7-22 Flat Roof Snow Load Versus Minimum Snow Load," STRUCTURE, November 2025.